#538: AI is Making Law Firms Obsolete, with Alistair Vigier

Episode Notes

Transcript

Zack Glaser (00:12): 

Hey y’all. I’m Zack. 

  

Stephanie Everett (00:13): 

And I’m Stephanie. And this is episode 538 of the Lawyerist Podcast, part of the Legal Talk Network. Today, Zack’s talking with Alistair Vigier about AI driven legal research and the lawsuit against his company case way. 

  

Zack Glaser (00:28): 

Today’s podcast is brought to you by iManage, so stick around and we’ll hear a little bit more from my conversation with them later on. So Stephanie, I am part of the Tennessee Bar Association pro bono subcommittee on enhancing A to J via technology, which is a mouthful, and I don’t know, do we have an acronym for it? So you got to kind of go through all that, but it gets me thinking or that gets me thinking every month at least about how we can solve or potentially solve or help solve access to justice issues with either everyday technology or by enhancing the technology that we’re building right now. 

  

Stephanie Everett (01:12): 

Yeah, there’s such opportunity here because I was just at a conference a couple of weeks ago and the numbers have not changed. I know lawyers are worried, but the amount of people in this country, especially, I mean I can speak for the US, that need legal help, that could benefit from legal services and that do not get them because they don’t know. They don’t know they have a legal issue or it’s cost prohibitive or they assume it’s cost prohibitive is huge. And that is, I mean, that’s just an opportunity, right? That’s a blue ocean. When you talk about red oceans being saturated, most law firms are trying to go after the same clientele at some level, and here’s an opportunity. There’s a lot of people underserved or not being served at all. And I think there’s opportunities for us to solve some of that with technology. 

  

Zack Glaser (02:10): 

And I think there are certainly opportunities. I think in this subcommittee specifically, it’s about pro bono, but there are opportunities in that either low bono space or like you’re saying, just places where people don’t know they even have a legal problem or it’s a problem that a lawyer specifically can solve. 

  

Stephanie Everett (02:35): 

Yeah, I did some work. I used to run an incubator for newer lawyers that we were teaching how to start socially conscious law firms. So I love challenging lawyers. How can we rethink the model? And here’s the key. When you talk a low bono, you can’t just say, I’m going to do the same thing that a traditional law firm does and just charge less. 

  

Zack Glaser (02:57): 

Right. 

  

Stephanie Everett (02:57): 

I mean, that doesn’t work from a business standpoint. You really have to think about, well, what’s the business model? How do you need to change the delivery of your services? If you’re going to charge less, you can’t just do the same. That should in sanity, that should be kind of obvious, but it sometimes isn’t. And so you can charge less, but then you’re going to rethink, is it going to be more of a limited scope offering? If so, what does that look like? Or do you take advantage and leverage technology? And there are companies out there doing this now. 

  

Zack Glaser (03:31): 

Yeah, so I think that’s one of the fascinating things that’s come about recently is if you look at the legal trends report, if you look at all these reports on what jobs artificial intelligence can do inside of a law firm, and it’s somewhere 40 to 60% depending on which you read, what a lot of what we do can be automated. And so we don’t necessarily have to limit our scope anymore to make things affordable to people. We can, but we can really, really automate things now with artificial intelligence, 

  

Stephanie Everett (04:13): 

Intelligence, but I’m going to challenge you a little bit. There could be an advantage, a market advantage to limiting your scope. So one of the products that I saw at the conference that I was featured, this woman is a lawyer, and she created a product that just does online, let’s see if I say this right, guardianship. She’s targeting moms and she’s saying how, Hey moms, you know how you sleep better at night? You figure out who’s going to take care of your kids if something happens to you. Yes, you could do that as part of your entire estate plan, but a lot of us are putting that off and we just really need to solve one little thing. Now, it could be a gateway to a bigger estate plan, but she’s offering a tech product at a really easy price to solve one specific problem because she can just do that all day long. 

  

Zack Glaser (05:04): 

And those people getting back to J issue, those people would not have done those guardianships. 

  

Stephanie Everett (05:11): 

Maybe they don’t feel like they need a will or they don’t understand they need a will, but they probably know if something happens to me, I’d like to be able to decide what happens to my kid instead of the court. 

  

Zack Glaser (05:21): 

Yeah, wouldn’t we? All right. Absolutely. 

  

Stephanie Everett (05:25): 

Yeah. So anyway, fascinating topic. We could probably sit here and jam on this all day. I can’t believe I just said it that way. It’s almost the holiday weekend. My brain is not working. But these are the kind of cool things that we love working with lawyers about because I, especially, this is where I think I have a lot of fun, is helping lawyers think about what are their services, how do they offer them differently? How do they maybe scope and price them differently? And then how can they use different tools to deliver the services? So I don’t know if you’re interested in that and wondering about it, let me know because love to work with you on it. 

  

Zack Glaser (06:06): 

Yeah, we love alternative fee structures and alternative structures here at Lawyerist, as you can imagine. Well now here is my conversation with our sponsor guest and then we’ll head into my conversation with Alistair. 

  

(06:20): 

Hey y’all, Zack, and I’ve got Tonya Wingfield from iManage here with me. And as you can imagine with the document management system, we are talking change management and how you get, I guess people and all your stuff on board in some of those knowledge management tools. Tonya, thanks for being with me again. 

  

Tonya Wingfield (06:40): 

Thank you for having me, Zack. 

  

Zack Glaser (06:41): 

So last time we talked about scoping the project in this change management, and this time, let’s say we’ve got all of our requirements and what we need to do done. How do we take care of the human aspect of the change management? How do we get everybody on board? 

  

Tonya Wingfield (07:04): 

That’s very important about that, Zack, because it’s easy for us to say we’re going to make change. 

  

(07:11): 

The hard part is making sure that the people we’re going to impact are ready for the change and ready to accept the change. And so that’s what doing a stakeholder analysis is going to help us determine. So with that process, we are going to actually get, a lot of times we can get this from your project managers and you may have to go out and solicit information from the user community, but we’re going to identify different people in their roles throughout the project lifecycle. So for example, the most important person who’s always the one writing the checks. So we want to always want to find out who’s the sponsor because they’re key. They’re the ones we want to keep satisfied 

  

Zack Glaser (07:55): 

During, keep them happy. 

  

Tonya Wingfield (07:56): 

Exactly. Never have them wondering or having to ask the question why. We always want to make sure we give them that. Those answers I have 

  

Zack Glaser (08:07): 

Preemptively. 

  

Tonya Wingfield (08:08): 

Great. Exactly. And so we want to look at our sponsors, and there’s usually two types of sponsors. You will have that initiating sponsor. Again, that’s the check writer. That’s the one that can have the most influence on that project. And then we have our sustaining sponsor, and this is often missed by customers. So this is after the project is done and everybody’s celebrated. We’re high, we got over the finish line. Well, who’s going to take care of this afterwards? Yeah, 

  

Zack Glaser (08:39): 

Yeah. 

  

Tonya Wingfield (08:40): 

That’s often missed. So we often call that the project owner or the product owner. And when you’re in a cloud environment, you want to remember your providers will do refreshes of the software, and that can happen in two week intervals. It can happen in a monthly interval, whatever that interval is. You want to make sure that somebody’s on top of those changes. Those changes are usually user facing. And I had a quick example. I was training on a project I trained on Thursday, and we got through Thursday, we came in Friday, the entire interface had changed because the provider did a refresh and we’re all sitting here, what’s this? What’s that? 

  

Zack Glaser (09:28): 

Where’s this button? 

  

Tonya Wingfield (09:30): 

And that’s why that product owner or that project owner is so important. They will decide how will we handle maintenance so that we can get this information out to the users and there’s no surprises. 

  

Zack Glaser (09:42): 

Okay, that 

  

Tonya Wingfield (09:43): 

Makes sense. So those are two important ones. Then you want to look at your technical team. 

  

(09:47): 

That’s usually going to be, if you’re doing it external, that’s usually going to be a vendor, but it could be also internal. So that would be who’s going to build your desktop and test it. If you have data that needs to be migrated, who’s going to take care of that migration strategy? So you want to make sure that you identify those individuals. They’re sometimes taken, are left off that list because it’s obvious they’re going to be part of the team, but you still want to make sure that there’s a communication plan on how information flows between the team and then your other stakeholders, like your sponsor. 

  

(10:25): 

Then you look at your targets. Your targets is your user community, 

  

Zack Glaser (10:31): 

People that are going to be actually in the trenches getting this thing. Yeah, okay. 

  

Tonya Wingfield (10:35): 

And they’re more important because you have a makeup of those. You have your resistors, you have what we call your advocates or your champions that are just have been waiting. 

  

Zack Glaser (10:47): 

They’re important. 

  

Tonya Wingfield (10:49): 

Exactly. Your champions are going to make sure that they share with their colleagues. And it’s good because it’s a peer to peer communication and they’re saying, look, I’ve been there. You’re going to love it. You’re going to always have your resistors. And so your champions or your advocates can help you win those resistors. And that’s another reason why that first step that we talked about last month, doing that scoping is so important. Now looking at all of those requirements and matching it, how it relates to these users. 

  

Zack Glaser (11:24): 

Users, gotcha. Okay. So we’ve got our requirements from last week. We’ve figured out what’s needed. We brought these people in determining what’s needed, and now we have to figure out, okay, how do we communicate what’s being changed to them and who is communicating those changes? Well, when you put it that way, it starts to make some sense, but when you just start going, how do I implement new document management or any new software in a large scale, that’s a big hill to climb and it’s a confusing place to be. I think for a lot of people 

  

Tonya Wingfield (12:05): 

It is. Especially when you love those SharePoint file where I have just a right click and it’s there. 

  

Zack Glaser (12:13): 

Absolutely. Absolutely. Well, Tonya, I appreciate you kind of walking us through some of this, and if people, this is a thing that iManage does a lot of, and again, if people aren’t familiar with iManage, which they probably should be because they’re out there and getting a lot of stuff done, but if people want to know more about iManage, they can always go to imanage.com. 

  

Tonya Wingfield (12:39): 

Correct. 

  

Zack Glaser (12:39): 

And they can connect with you guys. How else could they connect with you, Tonya? 

  

Tonya Wingfield (12:42): 

Well, from imanage.com, we usually have our site that’s set up and you have different areas that you can go to. So I would say mostly go look for success hubs that we have out there, and that will give you contact information on how to reach the actual individuals that are maintaining that information. 

  

Zack Glaser (13:04): 

Oh, gotcha. Fantastic. Alright, well Tonya, once again, I appreciate you sharing ledge with us. Thank you. 

  

Tonya Wingfield (13:10): 

Thank you, Zack for having me. 

  

Alistair Vigier (13:12): 

Hi, my name is Alistair Vigier, I’m the CEO of Caseway. A bit of background about me. I was in the military for seven years and I got shot and medically released the military, was willing to pay for my schooling. So I went to law school because I didn’t really know what I wanted to do with my life. Did well in law school but didn’t really love reading case law endlessly going through a hundred pages of decisions. So I actually never got called to the bar. I came back to Canada from the UK and I worked for a law firm for five years, and I became the business development manager and also a legal consultant. And I grew this firm’s revenue to 5 million in annual revenue, which is pretty good for a small boutique family law firm. 

  

(13:53): 

But I found myself getting pulled more towards legal tech than wanting to join the bar. And I connected with Ryan Gore, the co-founder of Clio in 2019 at Cio Con. And I asked him, if you were me wanting to get into legal tech, what would you do? And we kind of just became friends and started throwing ideas back and forward. And we started basically an AVO in Canada and the uk, which was a lawyer review website, and we became the largest lawyer review website in Canada and the uk. We grew it to over 200,000 people using the website every month over 25,000 lawyer reviews. But what we found was that people really, or lawyers, really didn’t love negative reviews and there was a lot of drama. And I think I got 20 cease and desist a week or something like that, which was just a little bit stressful, but no lawsuits, funny enough, just cease and desist endlessly. And the public loved it because there was transparency. They could check out a lawyer before going forward. And lots of lawyers got really positive reviews too. So there were lawyers that loved it, of course, but then Chat GBT came out in January, 2023 and all of the investment went towards AI B2B SaaS and we couldn’t raise money for a two-sided non-AI marketplace. 

  

(15:17): 

So that led me towards case way, which I can talk more about whenever. 

  

Zack Glaser (15:22): 

Yeah, yeah, yeah, that’s interesting because an interesting way to jump into legal tech and I think there are a lot of people out there that would love to have gotten their degree and then moved into the business side or the legal tech side, and there are a lot of people that would want to do that. But yeah, let’s talk about case way because as much as you got cease and desist letters in your previous organization, you’ve also gotten one here with the new company. But before we get into that, tell me what is it doing? 

  

Alistair Vigier (16:03): 

Yeah, so I’ll tell you what it is today and I’ll tell you what it can be. So case way today is it’s an AI that goes through millions of court decisions and it can answer questions, but the database is only court decisions. So unlike tragedy, GBT, which pulls from blogs, pulls from Reddit, pulls from Facebook, pulls from anywhere. If a judge didn’t write it, our AI doesn’t consider it. It’s all it knows. And it uses that for reviewing contracts, it uses it for creating contracts, for answering legal questions. And the number one issue that lawyers have with AI is hallucinations because they don’t want to be disciplined for something that’s just fake. So what we did was we started providing all of our sources and all of those sources were official court documents. And the idea was that if our AI says something, you click a link, it pulls you to where it says it in the court decision, and if what the AI says matches out with what the judge says, you can copy and paste it and you can use it right away and that it was the genesis of what it is today, but we’re going to be building a lot more than that in the future. 

  

Zack Glaser (17:07): 

And so, okay, that fundamental idea of you ask a question or you put a prompt into case way and when it gives you an answer back or it gives you some sort of relevant data back, it also says, this is where I got that information from. It cites its sources. 

  

Alistair Vigier (17:29): 

Yes. And they’re credible forces too. And it’s also trained to, if it doesn’t know the answer, it just tells you I don’t know the answer, which is another problem with hallucinations is chatt always wants to give you an answer even if it doesn’t know. And that’s where it makes stuff up. And it’s something that builds trust because lawyers are skeptical of ai, there’s no doubt about it. And it’s our job as a company to deal with that skepticism and build something that they can actually use and that they actually will use 

  

Zack Glaser (17:57): 

Well, okay, so in doing this, creating this I guess, legal research, artificial intelligence driven legal research platform, there has to be a lot of information back there. There has to be a lot of info that is driving this. And since you’re saying we’re not talking about just the internet at large, we’re not going into Google, we’re not just going and scraping all these websites, or I guess as Google does, it really organizes the websites for you to be able to search. But how do we as attorneys get that data? How do we have access to that data? How do we know that this platform is using good, relevant and up-to-date data? 

  

Alistair Vigier (18:48): 

So in the states, there’s 3,104 different databases of information, and I believe that’s every single county in the States has their own kind of database. Now that is insanely complicated. Now, this is where LexiNexis and Thomson Reuters made their money was they figured out how to combine all these, they put expensive AI products on top of it or other legal research products and they made billions of dollars from it. And that’s fine if you’re a firm that can afford these kind of products, but a lot of sole practitioners and small firms can’t afford these kind of products. Now, to answer your question directly, when our AI returns a result, we’re going to give you the link. So let’s just say, because California is the market we’re going into next, we’re going there next month. So I know more about California than the rest of the states. But 

  

(19:36): 

Let’s just say that you were looking up a case in Los Angeles and you’re typing in find me three cases where someone was found not guilty of bank fraud in Los Angeles. Between these years, whatever, super specific RAI will respond. And then on the side there will be the sources. Now when you click that, it’ll bring you to the official court document where the judge says whatever they said, there’s a button for summarizing it, there’s a button for a bunch of other things. So you’re good to go once you read the official court document and it says what the AI says. 

  

Zack Glaser (20:07): 

Yeah. So this actually, this makes me think of two cases out there. One, as I’m sure that a lot of our listeners knows is going to be West Westlaw versus Ross Intelligence, 

  

(20:19): 

Where Westlaw sued the bejesus out of Ross intelligence. And it I think may still technically be a going concern, but it stopped being a usable product or being a product that is being sold a couple of years ago, and Westlaw’s claim was that Ross intelligence scraped the data, not only the underlying case law, but one of the things that Westlaw does really well, and I think there’s just no doubt about this, is the key sites. Key sites are amazing. And I think that in that case, more than anything was you using the key sites to train your information and then giving back an answer as opposed to giving back a citation or a case that says, Hey, this is the type of thing that stands for this. And so that’s kind of one side of this type of thing of going and getting data that is case law, because case law, I think a fundamental thing to say here is that case law, the information underlying case law, the information underlying our laws is supposed to be free and open. That what happened in a case, what a judge said, I mean one of the fundamental things of our society is that we can get to our laws. 

  

(21:44): 

So on the other side of this idea is authors Guild versus Google, and that is basically the reason Google can serve up images to you and other things. But the reason that Google can serve up images and data to you in its searches is really because it’s saying, we’re not, this isn’t our thing. We’re leading you to this in a way. And so where does case way kind of sit in that area? And I’ve got a couple questions on top of that as well. This is fascinating to me. 

  

Alistair Vigier (22:22): 

Yeah. Well, I’m glad you mentioned Google because back in 1995 when the worldwide web was coming out, there were tons of lawsuits around indexing and people didn’t want Google or Bing to go on their websites and put them on Google’s website. They thought they were stealing images, stealing their traffic and all that. And it’s funny that now today people spend large amounts of money on SEO and it’s the exact opposite, but people get so scared with this stuff. And to answer your question about Way versus Kaley, which is a nonprofit, it’s not Thomson Reuters, but it’s the same thing. They were afraid that we were using their proprietary stuff to train RAI. They have practice guides, they have formatting and stuff like that. And I reached out to them and I said, look, I am willing to make our entire database available to you, and you’re not going to see a single piece of work that you did on top of the raw court data that comes from the judge in our database. 

  

(23:20): 

It just doesn’t exist. And the reason for that is that AI doesn’t care about formatting, it doesn’t care about any of that stuff. All it needs, at least our ai, all it needs is the raw words that the judge uses. Now, I don’t know, obviously the insides about how Ross AI worked and if they were using practice guides or anything that was proprietary, but I was super comfortable, and this is why I spoke out in the media, the lawsuit was like, I’m willing to be a hundred percent transparent to even make our database available, and this week we’re going to be making our database available to the public because otherwise you just end up in this. He said, she said situation where it’s like they say, Hey, case wave’s using our proprietary stuff. I’m like, no, we’re not using proprietary stuff, we’re just using raw data. I mean, who knows? And that goes on for so many years and stuff. So if we’re not using it, why don’t we just put it out there and then they can just check it themselves? 

  

Zack Glaser (24:11): 

Yeah, so I think that’s a good point. Which brings me to the idea of if our artificial intelligence tool and just kind like an artificial intelligence tool broadly as we use them in our law firms, if our tool is only going and looking at this kind of flat one dimensional raw data, so it can be case law, it can be our contracts, it can be information about our clients, things like that one dimensional kind of raw data, is it really helping us, it need some sort of feedback on top of that in order to indicate kind of relevance. This is why I think about the key sites in Westlaw would be, let’s take Ross Intelligence out of this thing, but the key sites in Westlaw for westlaw’s purposes are extremely beneficial because they give some context and some relevancy to that AI product, 

  

Alistair Vigier (25:15): 

Right? Well, I’ll tell you how it works with our product. So the first thing that we do in the states is we’re setting it up right now is that we’re setting up web crawlers. We’re not scraping. And I think that’s something that there’s a big difference between scraping and indexing as well. 

  

(25:28): 

Scraping is when you go to a website, you’re downloading 3 million whatever data points and you’re pulling it all in. And then whatever indexing like Google and Bing does, it’s, it’s checking your website, it’s pulling in information. As new information becomes available, scraping can crash. A website server indexing is very gentle and it’s not invasive whatsoever. Now the next thing that we do on top of that is we clean the data with an algorithm because if we’re pulling in information from 3,104 different data sets, we’re cleaning it for our AI and RAI only wants the raw text 

  

(26:02): 

Data that’s at least RAI. And then on top of that, we’re writing algorithms for search to improve the search. So for example, when we started, there was a law firm that was trying something out and they wanted to find this super specific case that they were aware of on case way. And it was just something like find me a case where a Crown prosecutor was assaulted in Vancouver and the person, the defendant was found guilty and case way being honest said, I can’t find the case. And then the law firm was like, Alistair, this sucks. It doesn’t work. So I went in and looked manually at the case and the judge used the term charges are dismissed instead of not guilty. So we had to train our algorithm. Those mean the same thing. So there’s a lot of that on top of it. So there’s different steps that go into all of this 

  

(26:51): 

That make it much better. But to answer your question, how does the AI get better? You’ve seen it on chatt BT where there’s the thumb up and there’s a thumb down and then there’s AB testing on both sides. Which response do you like the most? That is the easiest way to train an AI because it really doesn’t matter what I say or what anyone else says, it’s the lawyer who’s actually using this or the paralegal or the self-represented person. Was this result useful for you? And our team will manually go into something where we’re constantly getting thumbs down. Let’s just say maybe it’s immigration law for some reason there’s something wrong going on. We investigate it, then we rewrite our algorithms to improve it. If we’re getting a 90 to a hundred percent thumbs up rate, we don’t have to worry about it. That is how we train our ai. It’s based on the user feedback. It’s not based on citations or any of that kind of stuff. 

  

Zack Glaser (27:41): 

Okay. So the case way then theoretically is getting better as it goes. It’s learning on our input as it’s coming back. It’s figuring out what is more a cat as it goes, right? 

  

Alistair Vigier (28:02): 

Yeah. And not only what I just mentioned before with the thumbs of the thumbs down, the great and terrible thing about lawyers is that they’re super willing to provide their feedback and sometimes you don’t want that. They can be quite harsh about it. But other times we have a lawyer that’s like they write all these screenshots almost like in an affidavit format and they’re like, I tried this. This was the result I got, this is what I wanted. I tried this, blah, blah, blah, blah, blah. For us though, as a tech startup, we love that because we’re getting so much feedback. I think we get 40 lawyers email us every single day about feedback and they’re telling us what they love, they’re telling us what they don’t love, and then they’re having suggestions. For example, lawyers love Word. That is something that we were originally going to try and get lawyers away from. 

  

(28:46): 

Why don’t you stay on our platform? Why don’t you use Google Docs? More tech friendly things. Every lawyer we talk to, they’re like, I want to be able to download these results in a Word document. So we just give it to them. And that’s something that I wouldn’t think of. It’s certainly not something any of our AI engineers would think of, but that feedback that comes in every single day makes our product so much better. And Mark Brit, the founder of avo, I was down in Ccle con talking and talking to him and he said, meet the lawyers where they’re at, not where you want them to be. And that’s something that really stuck with me is that we could build all this really advanced ai and the lawyers would hate it. They wouldn’t use it. They’d be turned off by it. They want a Word document, just give them the Word document. They’ll be happy. They’ll sign up, they’ll pay. 

  

Zack Glaser (29:33): 

Yeah. I think that’s a tough thing to a admit is you’re in innovative tech space. So it’s a tough thing to admit of like, okay, we are going to use Word because personally I prefer Google Docs. Absolutely, a hundred percent. But the other side is we do want to be dragging people forward a little bit. Maybe not kicking and screaming, but we do want to be moving people forward a little bit and getting them used to this. So talk to me then about what you see for Case way in the near future as we bring people along kicking and screaming. 

  

Alistair Vigier (30:19): 

So I think the easiest thing that we can do is something called a bespoke agent. Now the bespoke agent is where a law firm populates our AI database with their own forms, documents, and contracts exactly the way they want to do it. And that doesn’t require us to collect case law because with the case law, I foresee tons of lawsuits. It’s not going to end with this ly. When we go down to the states, I think at some point Thomson Reuters or Alexis Nexus, even if we’ve done nothing wrong, we’ve never accessed their site, 

  

(30:48): 

They’ll be threatened by us. They’re going to come after us. I just believe that. Or maybe we’ll do a partnership with them. But the idea that these organizations are going to allow someone onto their turf and they’re just going to be like, oh, well they’re not really doing anything wrong. Let’s just let them be. I just don’t think it exists. But we’ve talked to a lot of law firms and they know just for example, a pageant law firm I was talking to at CLE Con, they have 30 attorneys and they know exactly what they want at first draft of a patent application to look like. Now what they’ll be able to do is they’ll be able to populate our database with that first draft. They will then input all their client’s facts into it and it will create the first draft of a patent application for them within seconds. Something that would normally take a junior lawyer or a paralegal a week or two weeks to do 

  

(31:32): 

Check. It’ll do it, and then their lawyers or paralegals will take it from there, check everything, perfect it and prove it. It will work the same way with business law firm creating a shareholder agreement, a family law firm creating a divorce application. And the nice thing is that they are populating with their own data and there is no agreement upon law firms on exactly the right way to do a certain document so they can have it exactly the way they want. The other thing I’m really excited about is, and we’re actually talking to Fin Pay, which is the My case group 

  

(32:05): 

In Texas, about a partnership they want us to build, I can’t say too much, but I’ll say what I can. They want us to build a form upload for the courts. So for example, it takes law firms a lot of time to upload something to create these forms for the courts. And there’s all these memes out there about the court clerks who are harsh and they rejected, and one court clerk likes it this way, the other court clerk likes it this way. That is a space that they’re really interested in. And that’s something that we can also do with our bespoke agent where the firms will upload how to start a lawsuit, the forum in that county for that, they input the client facts, it will shoot it out the exact way they’ve done it a thousand times in the past, and then that saves them a ton of time. And again, that doesn’t require anything to do with the case law. So we’re still moving forward with the case law, but I think there are lower hanging fruits which will enable us to focus on innovation and less on lawsuits. And so that’s where I really see this going. 

  

Zack Glaser (33:08): 

So that kind of keeps you going into the ape thing that you’re talking about here. Keeps you going into some of these individual counties. Then in a way, I think you and I talked a couple months ago about trying to enter into the United States space, and I’m an attorney in Tennessee. We have 95 counties and 96 or 97 general sessions courts and 96 or 97 general sessions court clerks. And even though we have an administrative office of the courts that puts out specific forms that are supposed to be uniform forms, some of the bigger counties still don’t use those. And so how do you see being able to get into a conversation? I guess with the counties like that, 

  

Alistair Vigier (34:05): 

It all starts with building it first. If we reach out to the counties first and we’re like, Hey, we have this idea of how we’re going to revolutionize the courts, they’re going to be like, okay, cool. We’re not interested in talking because we’re busy. If we have a working prototype and we can get a few law firms behind that working prototype and then we reach out to the courts, that’s where it gets exciting. If we can get a few legal influencers that have been talking to in California behind this, that makes it possible too. So we kind of need to build it, we need to demo it for them, we need to get in front of here. We call ’em the chief justices, they’re the decision makers, and then we need to generate buzz around it. So once we have the court form update, we are going to try and get it out in the media and say, it’s ridiculous that this is the way that courts are currently doing it. There’s a way better way, let’s partner. Let’s make law more accessible, let’s improve that, the access to justice problem, let’s make it easier to practice law. And then it’s just a process of going county through county trying to get the state bars involved as well. I know that some of that Clio did a really great job on was partnering with the state bars and just trying to get as many people behind us as possible. And certainly not going to be a fast process. 

  

(35:22): 

If we can partner with organizations like Clio and Afinipay, that makes it easier because they built a lot of these relationships over time as well. And that’s what I think Case Way is really amazing at doing is that we’re a small team, but we punch way above our weight. We build really amazing products in Canada. At least we have built the best legal AI products, in my opinion, for the cheapest price, $49 a month. And that is really impressive, and that’s how we’re going to approach this in the states too, where we’re going to build really amazing AI products. We’re going to be out there talking about it and we’re going to make it affordable. And then we’re going to be patient too, because that’s the problem. A lot of tech people who come to the legal space are or who don’t have a background in law is that they’re like talking to the lawyers. They’re like, oh, get in. It’s super amazing, fast, exciting. And the lawyer’s like, I hate that. I want to be the last person I want this to be. Totally. So there’s always going to be patience. It’s not going to be like, let’s do this. 

  

Zack Glaser (36:21): 

Yeah, let’s talk about that. Let’s talk about a how long it kind of takes attorneys to adopt these things, but more importantly, thinking about the specific scenario with Case way as it comes into the United States. You’re saying, and I believe you’re correct, inevitably there’s going to be a lot of lawsuits here, and so there’s going to be one front, at the very least, that is a pretty large slog for you guys. How does an attorney go about saying, okay, well, I’m still going to get into that platform. I’m still going to try to use that platform even though, and no matter what the risk is, even though there’s a risk that it somehow goes the way of Ross. So now I’ve invested my time, I’ve invested my knowledge into that platform. Why would I do that if it’s not tried and true? 

  

Alistair Vigier (37:19): 

Yep. Yeah, that’s a great point. So the first thing that we said right after the lawsuit is we had a thousand people sign up a thousand lawyers because they wanted to try it. They were curious about why this nonprofit was coming after a startup. That was the biggest problem for us before the lawsuit is how are you relevant in a space where there’s so much noise, there’s so much talk and stuff like that. We became the most relevant AI, legal AI company in Canada basically overnight. Now, that was absolutely the conversation I had with people or law firms was like, do we want to invest our time and money into this if you’re just going to stop existing? So three things. I said, number one, just do month to month, we have annual rates. If you’re paying month to month, there’s really no risk. Your risk is $49 for this month. 

  

(38:03): 

The other thing is that no infrastructure tie in right now for Case way. It is just a legal platform, so you don’t have to tear out your infrastructure. There’s no setup fees. It’s just you use this as an external web platform that’s on the cloud, so it doesn’t matter. The third thing, and this is very different, is that I am not going to go away at all. If I had to reincorporate in The Bahamas, if I had to, I don’t think I’ll, but if I had to be in a beach, if I had to be in Africa, if I had to do this from Shanghai, I would do it. We are not going to go anywhere. I’m absolutely a fighter when it comes to this stuff, and I don’t care what any organization does. I firmly believe in this. I believe it’s good. And we get emails from testimonials all the time from lawyers and they’re like, this is an amazing product. Thank you so much for building this, and that gives me the conviction. So we will never shut down. I don’t care. Even if the lawyer’s like you owed, it’s so much better if you shut down for a few years, we’re not going to do it. I don’t care. I’ll live on a beach in The Bahamas and I will operate this, not just Come on me. 

  

Zack Glaser (39:12): 

I love this because it actually gets into something that I think is a broader topic that I have trouble finding people to talk to about this. They don’t always take me seriously. I didn’t expect to talk about this. 

  

(39:27): 

We have to innovate as a community because we can only regulate so far, literally physically, we can only regulate within the bounds of, for our purposes right here within the bounds of North America or within the bounds of North America and the places that we have treaties with. But if I say, I have access to American law and I can go get this, and I could then spin this company up in someplace that doesn’t necessarily have a treaty with the United States and extradition issues or anything like that. And I obviously hope that that is not going to be the case with Case way. Me too. 

  

(40:13): 

But even that is a thing we have to think about as innovators, as regulators, as people who are dealing with these things, is that we are operating in a place that is not completely, is not completely controllable. I don’t know what my question is with that other than highlighting it that if I wanted to right now, I could go move to another country and spin up a company and an organization and a website that practices law in the United States that files divorces in the United States. A hundred percent could. And you can’t do anything about that. And that’s the thing that I think is crazy when people just try to push back on these things and say, well, we can’t have non-lawyers owning law firms. We can’t, can’t be using artificial intelligence in our, we have to completely regulate the use of artificial intelligence in the law practice. We can’t regulate everybody. So I think that’s an interesting, I want to make it a tangent, we could talk about this for a while, but I think that’s an interesting tangent, is that you can’t just say, okay, well this has to shut down. It may have to move. 

  

Alistair Vigier (41:32): 

Right? Well, it’s harder for you because you’re a practicing attorney as well, so they always have your license to threaten you with, so you can’t just do whatever you want, but for me, who doesn’t have a license, I don’t care. And you could imagine setting up in a country and just like, I’m not going to do this, but an organization could scrape LexiNexis, Thomson Reuters. They could put out a product for really cheap and totally mess ’em over maybe in Cuba or something, I don’t know, in China. And that’s what I also told Kaley was like, this is me today at my most collaborative. I am trying to work with these organizations, but when they say we’re not going to do an API when we’re not going to make our data available, when we’re going to so cute when they just don’t want to work with AI companies, it’s like they will move to wherever they have to move, and then they’ll just do whatever they want and it’ll be much more destructive to you. And that’s why a lot of law firms, after I said this, I think they felt my passion and they were like, okay, case way is not going anywhere. This guy, he’s half right half insane maybe, but he’s not going to shut this thing down. 

  

Zack Glaser (42:45): 

And I think that’s important to say, but it’s also important to understand that it’s possible that is a reasonable thing to be able to do a reasonable reaction to that. And I think that’s a good point too, is that in this scenario, you have an organization that is saying, listen, there is this stuff, this underlying data that is we have already said is not copyrightable. It is the law. A fundamental tenant of our society is that you can’t hoard the law. Okay, well, I’m just using that, and this is what you’re saying here is I’m just using that and I’ll show you that I’m just using that and if we can’t be okay with that. But yeah, I think that’s important to note that there has to be kind of reasonableness both directions here. 

  

Alistair Vigier (43:43): 

Yeah. Yeah, a hundred percent. And it’s really hard to change the law when you’re in law. And I think that’s why lawyers have a hard time innovating is there’s always that risk. And you’re so caught up in what the law society is saying. The Bar Association is saying what the courts are saying, and there is a different way of doing it where it’s like, I can see the future of law and it’s absolutely going to be ai. There’s no doubt that’s true. I want to work with everyone. But yeah, you said it perfectly. I am willing to be reasonable at this point, but if I feel that I’m just being attacked regardless of what I do, then I’m just going to do it my way. 

  

Zack Glaser (44:27): 

Well, yeah, and I think it’s important. So I talked with Matthew Butrick a year or so ago about a case that he is a part of that is against copilot and some other organizations that we’re scraping we’re GitHub and Deviant Art in order to create platforms that competed with the people who had put the information out there. And we were talking about fair and ethical use of data in ai. I think it’s important to note that that information had a human creation aspect of Yes, 

  

Alistair Vigier (45:09): 

Yes. 

  

Zack Glaser (45:10): 

We’re talking about underlying information that is copyrightable in the first place in that matter. And so if one were to say, well, I’m going to scrape this whether I’m, I’m going to gather this information and I’m going to use this information whether you like it or not, that’s an unfair and unethical use of data. Agree. I think that’s really clear. And so I want to make sure that we’re differentiating this where the underlying information that you’re saying that case way is getting here is case law. 

  

Alistair Vigier (45:44): 

It is. 

  

Zack Glaser (45:45): 

And as you highlighted just the information you’ve taken, the formatting and anything like that out of it. And then the second step of case way, the second product is being able to leverage artificial intelligence against our own data, against our own things inside. How are we going to have, let’s talk about that. Let’s pivot to that just a little bit. How are we as attorneys going to have enough data to really get relevant information out of that? 

  

Alistair Vigier (46:23): 

Most law firms that I’ve spoken to have a massive amount of data. What kind of law did you do or are doing in Tennessee? 

  

Zack Glaser (46:30): 

Yeah, so I did collections. I had a massive amount of data. 

  

Alistair Vigier (46:35): 

So you send out collection notices, maybe lawsuits. 

  

Zack Glaser (46:39): 

Oh yeah. 

  

Alistair Vigier (46:39): 

Cease and desist letters. Those are probably the key documents you have. I assume that you have hundreds of each of those, or thousand, 

  

Zack Glaser (46:49): 

Yeah. Yeah. You mean hundreds of templates. 

  

Alistair Vigier (46:52): 

Each of those hundreds of templates? Examples, you sent it to this person you sent to that maybe there was some small differences in terms of where the state was that you were sending to. Maybe one was for credit card, maybe one was for, I don’t know, buying lumber or something like that. The point is, the more documents you put into it, the better it’ll be. But even if you only had a hundred examples, it’s going to be able to create it with the words you use. It’s going to be able to create it in the format you use. Do you like paragraph headers? Do you not like that? What kind of font do you use? Are you really inflammatory in how you talk or is it very collaborative just saying, Hey, I understand you might have tough times and I just want to work out a deal here that might be different how you practice versus someone else. 

  

(47:39): 

The point is that obviously cease and assist probably don’t take you an insane amount of time, but if it took you an hour, it would populate this for you within a few seconds. You’re just going to go through it, make a few changes, check it and send it out. And obviously the more laborious the document is like a patent application, which is highly technical, it adds even more value because that can take someone a week or two to create or a notice of civil claim or something like that. So that’s how it works in that regard. 

  

Zack Glaser (48:10): 

Okay. So yeah, that kind of helps me shift my mindset a little bit from, it’s not so much about the creation of the underlying information within that letter, the law of that letter or something like that. It’s about making that letter out of what you already have, making it look like it’s from you, making it seem like it was written from your office. And I think that’s a little bit of a distinction that is important because in that scenario, we’re not so much mining the data, mining these letters in order to find an answer to a question mining and answer to the question as we would think of it, we’re mining these letters to find out how would you word this? Where would you put the signature line? How would you format this? Would you italicize titles or would you underline them or something like that. 

  

Alistair Vigier (49:10): 

Yeah, yeah, exactly. It’s much easier for us because we don’t have to worry about the data collection. With the data collection, I have three categories. So the green zone, the gray zone, and the black zone. So the green zone is, it’s publicly available, non-pay walled, no copyright, and we should be able to use it. Everything that we’ve done to date is green zone. The gray zone is something that open AI has been alleged of violating where they’re going to things like the New York Times journalists collecting stuff that is copyrightable and possibly paywall and that sort of thing. And then the black zone is like, it’s not even on the internet. You’d have to hack into a system to collect it, and that’s obviously criminal. So there’s three different zones on that. But when it comes to the bespoke agent, it’s a law firms that are populating it with their own documents. 

  

(49:59): 

And the other thing that we’re going to offer is do you want to have it saved into memory or not? And some law firms I’ve talked to, they want it saved into memory so that they don’t have to keep populating our database with the documents, but some of them just want it deleted after every session. And that’s for confidentiality to protect it, to make sure the AI isn’t trained on it. And that’s something that we’re working on right now with the case law searches as well as our default is everything’s deleted after every session, but soon we’re going to have it so that you can pick if it goes into memory or not. Because some things, you might just be doing research, so there’s nothing confidential, but if you’re uploading something with a client’s name or client facts, you might toggle it to no memory for that period and then go back to the memory. So I think that’s another thing that lawyers are really excited about because they told me is the ability to switch between the two. And I don’t believe that exists in other AI products. That’s what I’ve been told. 

  

Zack Glaser (50:53): 

Yeah. I don’t know that it does or doesn’t, but I can easily envision that scenario where in one case I really want to have this prompt be detailed, very detailed, prompt. I’m working with case ways like a thought partner and we’re working through this thing. And in another case I’m saying find a case, find me a case that is specific to this type of thing, and it’s like, I’m going to ask that a couple times, or it really doesn’t matter. I don’t want any client data in here. And so we can go back into the system added fin item as much as it wants. 

  

Alistair Vigier (51:32): 

You can also upload your work products. Even today, you could upload your draft of a notice of civil claim and say, pretend you the judge and use case law to find problems with this. And it does a really amazing job of that. You can role play with it, and it’s like you mentioned this, but in paragraph 59, you’re saying someone that’s inconsistent, you mentioned this, but you didn’t get specific with a date. It’s pulling out all these problems that judges have had in previous court rulings. 

  

Zack Glaser (52:01): 

Yeah. Okay. So I think that’s interesting to dig into for a second is that the idea that this isn’t just a search tool, it is something like in the vein of chat, GPT where you can continue to query against it, continue to refine, have almost a conversation with it, 

  

Alistair Vigier (52:27): 

Right? Yeah, exactly. And I use it for all of the contract review as well. Every single time I get a contract from a contract or someone like that, I put it through there and I’m like, pretend that you’re representing the company and as a lawyer and give me some information back about issues with this. And it says, this clause was found not enforceable in a recent court decision, so you probably don’t need that. This kind of situation is too broad. A judge has said that a non-compete, I’m just making something up. A non-compete that exists beyond two years in this jurisdiction is not enforceable. They’re asking for a non-compete for five years, stuff like that. So it can do a lot and it’s only going to get better. 

  

Zack Glaser (53:13): 

Yeah. Yeah. Okay. Well, so yeah, thinking into the little bit in the future, looking into your crystal ball, running out of time here a little bit, how do you see artificial intelligence shaping the way that law firms conduct legal research in the next, I’d say five to 10 years, but really one to three? 

  

Alistair Vigier (53:39): 

I’d say it’s already happening today, but they’re lawyers that are more willing to dive into the space than other sectors. 

  

(53:47): 

Today I’m seeing a lot of flat fee lawyers using this just because it makes so much sense. If they’re charging $3,000 for a shareholder agreement and they can get that done within one hour using ai, they’re making $3,000 an hour. I think it’s a little slower for lawyers who are charging by the hour and their model is baked into inefficiency. If it takes them 10 hours to find something and they’re charging $400 an hour, they make $4,000, do they really want to do it in an hour? That’s a harder conversation. We’re having, a lot of government lawyers use this, a lot of in-house counsel because if they can get something done in one hour that normally takes them a day, they can spend the rest of the day on Instagram and their boss doesn’t know, and that’s a sweet deal for them. Some lawyers are certainly moving faster than others at some point. 

  

(54:35): 

I think some AI company maybe me, will have to really start a storm where I think at some point it’ll be unethical for any lawyer who’s charging by the hour to not use ai. I would make that argument lightly today and more intensely in the future because again, if they can do something that takes them two seconds, it normally takes them 10 hours. Their client’s just paying for that, and that sucks. They shouldn’t be doing that. Now, it’s not me telling law firms, Hey, you’re being unethical. That’s not going to be well received. I think it has to be the state bars and the law societies that are going to get behind AI and really push that. But they have to get there themselves first. They need to have AI policies, they need to understand and they need to support ai. That’s probably going to take three years. 

  

Zack Glaser (55:22): 

And I think that’s a really good point is that when we say, I agree with you completely on the ethics of or professional responsibility of using artificial intelligence, I think we’re exacerbating this issue of do I want to use technology that’s going to speed me up if that just means that I’m going to literally make less on this? But I think when we say that we argue it lightly right now, it’s not so much we argue it lightly because we don’t think it’s correct. We’d argue it lightly because I can see where somebody would say, listen, I got to get my ducks in a row. And so I think the takeaway from that is that yes, it may not be unethical to not be using it literally today, but I would 100% argue that you need to get your butt moving in that direction. 

  

Alistair Vigier (56:17): 

Yes. 

  

Zack Glaser (56:19): 

Well, I think that’s a great thing to end on here. Alistair, I really appreciate your insight into all this and your candidness with talking about this. We could go into four different conversations coming off of this whole AI discussion, but yeah, if people want to, we’ve been talking about case way in a broad sense. This is a lot of artificial intelligence out there, but if people want to see Case way, obviously they can catch you over at Caseway. I think it’s ai. 

  

(56:55): 

So Case way.ai. Alistair, I really appreciate you being with me. Thank you. 

  

Alistair Vigier (56:59): 

Yeah, thanks so much for having me. 

 

 

Continue Reading