Darin Thompson

2Thanks for having me. I'm excited to be here and looking forward to talking to you about hopefully a range of interesting and maybe related, maybe unrelated topics. Let's give it a shot.

5Yeah, that's— those are all interesting questions. The first one that I'll talk about is, I guess, why I'm interested in doing things differently than maybe they seem to be done. Just generally, I think that the law is full of so many incredibly smart people, lawyers, judges, a range of different legal professionals, people working in and around the system. We got a lot of collective brain power. Like, it's I think we're really fortunate, really privileged in that way to have so many smart people and driven people too. Let's be clear, we— the, the classic, the cliché of the lawyer is the driven person, the A-type. Yet we still have a lot of problems, we have a lot of challenges, and For me, I was, I would just never settle for it.

6I thought that there were still a lot of different things we could do differently, a lot of ways we can improve the system and a lot of, a lot of things we could do to make the law work better for both for litigation and for transactional type types of law. So When I was practicing, I spent a short time in private practice. A partner in my law firm one day was talking to me at a Christmas party. Everybody had a few glasses of wine and was sort of letting down their guard, although this person who I just think is a great, great guy and a great lawyer didn't really need wine to let down his guard.

7He was pretty forward most times, but he said to me, He said, "What is it with you?" He used a little more colorful language than I've just used, but anyways, he said, "What is it with you? I send you out to do a research memo and you come into my office, you know, half a day later and say the system is doing this wrong and the courts are doing this wrong and the law treats this thing wrong." And I would say, "That's fine, just, you know, go out and finish the memo. I just need that memo." And A couple hours later, you'd come back in my office and tell me what you'd found and then be telling me all the things that were being done wrong.

8And I think he nailed it. He— I had trouble. I could, I could still write the memo. I could do a pretty good memo at the end of the day. But I would have a terrible time getting caught up on all of the things that were happening, all the ways that the law was being done that I couldn't understand why they would be designed that way and why we continued to do it that way, even once we realized that it wasn't working well. So that kind of brought my— that was kind of a good illustration of where my compass and my orientation, I guess, points me when it comes to law and how things are done.

9So as I got into— later on, I started working for government and I was fortunate to work in an office that was pushing hard on the justice system to try and find ways to improve the system. Lots of looking at alternative dispute resolution, lots of, you know, different ways to sort of push the public justice system in, you know, away from the bad and in a direction towards the good. And our office realized around 2008, 2009, that technology held a lot of potential promise. It was still sort of undiscovered territory at that time. There was some e-filing, British Columbia has been e-filing since 2005. And, you know, doing some court information, court documents online, those sorts of things, but we were really interested in ways that we could use technology to enable new processes.

10So since then, what I've come to realize is I am still very interested in technology, but I'm interested in technology second. What I'm really interested in is— maybe I'm interested in technology third. What I'm really interested in is users and what the public needs, the system, rather than the justice providers. I'm interested in the users. And secondly, I'm interested in creating the best processes we can. Amazingly great, insanely great, to borrow Steve Jobs's language, insanely great processes. It turns out that technology can be used to enable those. Technology is the tool that can support a lot of these processes. So I've come to a point now where I don't start out with, okay, how can I use technology in the system?

11My starting point is, what do my users need? Next is, how can we build some insanely great processes for them. And then third, what tools can we use to make these processes come alive, and what platforms can we use? And very often it is technology, but by no means exclusively technology. Humans are still at the core of a lot of the, the things that I'm interested in.

13Yeah, absolutely.

16Right.

18Yeah, I don't know if there's something about the West Coast. It's probably best to leave it for the people in Central Canada and the East Coast to decide what it is that's unique or strange about us here. On the West Coast. It is a straight shot. We're the same time zone as San Francisco and Silicon Valley and Stanford, where a lot of legal design happens. But I know there's, there's people interested in legal design all over the place, including the other side of North America and Europe. But it's interesting that you mentioned the two ways to frame this discussion. One, you mentioned focus on users, that's what I'm very passionate about. The other one is that, you know, how can we avoid touching paper?

19And I think that's, you know, I think you're getting to a really important way to, or two different ways to frame our discussion. On the one hand, like I said, we could create, we can look at creating processes that are value or outcomes for users. This is for the people that we serve, the members of the public, the businesses, anybody has to resort to the justice system to get some sort of outcome or output. So we can focus on them and design processes that work for them, design processes to deliver that value, but that still work for those people. Let's look at the other way of framing it again. How do we avoid touching paper?

20I think what I hear from that, or what I take from that, is quite often that is an effort to digitize the existing system. So when you talk about avoid touching paper, you're still talking about a document, or you're talking about a form— the F word for me when I'm doing legal design. The you're taking the artifacts of the current system and the steps we do in the current system, and you're trying to find ways to create electric electronic substitutes. And that's been very popular since the pandemic started, of course, when the courts effectively shut down for several months. They had to figure out how to get things moving again, and what it seemed to me was a lot of the efforts were focused on creating digital versions of the status quo.

21So let's do video hearings, you know, let's, let's find ways to exchange electronic documents. So I applaud those efforts. And I think, I think that's great. But what's the focus in that second type of framing? The focus is on our processes and on the things we use and on our forms and creating electronic versions of those. It's different, in my mind anyways, it's different than focusing on the users. I start with What does my user want? What do they need? What are they capable of? What is it that they can't do? And try and design a system around them. I'm putting them at the center, as you say, the user-centric design. And again, it's different than saying, okay, here's how we make the widgets and how we've made the widgets for the last few centuries.

22Let's see if we can create an electronic version of that process of making, making the widget. So I hope, I hope that kind of comes through. You can, you can see the differences there. To boil it down to a form, rather than create an electronic court form, I would, I'd boil things down to first principles if I could and say, what is it that we need from this user? Probably going to discover that we need some of their information. Great, let's focus on the information. Let's figure out what the best easiest way is to collect information from somebody. Is it going to be a court form? Maybe, but if we do user testing and take a user-centric design approach, I doubt it's going to be an electronic court form.

23It'll probably be another way to extract information from a user. And then once we get a hold of that information, let's put it to work in our, in our system. So those are just examples again, where you really boil things down to first principles, focus on the users. You're less hung up with the way things have always been done, and you're not just trying to create a digital version of the status quo. The last thing I'll say though, because I know I'm into dangerous territory here for lawyers, not just on the cultural side, but from, you know, the principles that we hold dear in justice, is it's entirely possible to address within this creating a new process sort of thing approach, it's entirely possible to do things like address procedural fairness and to create fair and, you know, just processes for handling these disputes.

24There isn't only one way to do it, the historical way.

30First, I should say too that the greater use of technology— I should be careful not to corner myself into a dichotomy here. I am not opposed to some of the changes that you've just talked about.

32no, no, no, no. This is something I suffer from. From myself. So I want to say that I do support this more incremental improvements using technology. Not opposed. If they're, if they're done right, then fantastic, bring it on. But I think, as I said right at the beginning, we're a, we're a profession that's full of like really, really smart people, and collectively I think we have the potential to aim way higher and do a lot more and make much bigger improvements. So in terms of the question that you asked, though, I hear this one a lot. When I first started talking about using more technology in justice processes, that was definitely something that people would bring up, like, aren't you gonna put people out of business?

33Put people out of a job, the people who work in the registries, and eventually judges, if you're, you know, having systems that can do decisions, you're trying to put judges out of work. The first thing I'll say is, from a business case perspective, yeah, people are the important— or sorry, not only the important, but also the most expensive part of our system. That is what makes our system more and more expensive year after year, is the people that we have to pay to administer our public system. But we have so much backlog. The unmet demand in Canada and most other developed countries, the unmet demand is so great that we are not, I think, going to be firing any people if we created more capacity in our system.

34If we create— if we increase the capacity in our system by an order of you know, say 50% more capacity, 75% more capacity, I think we're still not going to eliminate the unmet demand, all of the demand that the system currently can't address. So what I'm more interested in is saying, okay, we've got people working in the system from the Court of Appeal judges and justices right down to the clerk who works in the registry. What I want to do is design systems, design processes, design systems, use technologies that free those resources up, that help them climb up the value chain to do more interesting, more rewarding work that frankly can't be easily automated.

35So Another way to look at it is if I have a person who's dedicated to their job and they're smart and they care a lot about the system, they care about the meaning, you know, intertwined with their role, I don't want them to be doing a job that is mindless and repetitive and, you know, manual. I don't want them to be doing that all day. I want them to be doing something great. Instead of stamping a document over and over again. I want them to help, you know, help the litigant, the self-represented litigant who doesn't even know how to start. I want them to be talking to them or helping them or doing a more complex task, not doing the same repetitive thing over and over again.

36So this view, by the way, applies for legal practice as well. You know, according to some people, so much of what lawyers do is just repetitive, work that absolutely can be automated, we should automate that. Are lawyers going to be out of work? No, because there is so much demand on ways that we're not serving the public. Imagine if we could automate some work and drive our volumes way up and our margins down a little lower and find ways to serve a lot more people more economically. You wouldn't even be scratching the surface of the unmet. Demand. So yeah, I don't— I'm not, I'm not seeing us laying off people just because we're starting to use technology.

37What I see instead is freeing up people to do more important things, to make the best use of the human resources that we have, and having to hire new people, potentially, who know how to program the technology, who know how to, you know, build and maintain these systems. So I— we've taught in a few universities. We've taught students our methodology that we call knowledge engineering, where you're capturing knowledge from a human expert and putting it into an intelligent system so that that system, once you turn it on, could just serve people over and over again, 24/7. Well, that's a new role— a knowledge engineer who, who learns the methodology that we created and goes out and do this.

38So that's just one example of a new job. But yeah, anyways, I don't know, are some of those points resonating? Does it sound convincing to you?

40Yeah, the one that was built out here in British Columbia is called the Solution Explorer, and it's the first dispute resolution phase for people coming to the Civil Resolution Tribunal. Before I jump into what the roles and functions or how that system is used for the Civil Resolution Tribunal, I'll just say generally expert systems are It's a system, a technology-based system that's made to imitate or emulate the interactions you might have with a human expert. So the, the system is intelligent because we've gone out and collected knowledge from humans and put it into a knowledge base that's within a system. But the system has to be designed— the knowledge base has to be designed in a certain way so that the system can work with it.

41If you do it right, a system, an expert system like this, can— it can make deductions, it can it can apply conditional logic to certain inputs, it can apply from a— to get more into law, it can apply facts to law. So the user's putting in their facts or circumstances, the system can apply those to law. So To bring it back to the Solution Explorer that the Civil Resolution Tribunal uses, I usually organize it into 4 sort of buckets, and in terms of defining the functions or the value that it provides for citizens here in BC. Firstly, it can diagnose your dispute, and that's really important for people who want to engage with the justice system.

42If something bad happens to you in a transaction and you're a person who's not familiar with the law, you're not going to have an immediate sense of how to characterize it in legal terms, right? So there's nothing like organic about the way we come up with some of the causes of action or the statutes that might apply. These were all invented by humans, and only the people who have the knowledge know how to do it. So we try and deliver some of that through the Solution Explorer. So it can help diagnose a dispute as a, you know, either a consumer dispute as opposed to a motor vehicle personal injury dispute. And if you are looking into the consumer basket or domain, you can diagnose it further as something to which consumer legislation applies, because you know that this person bought something from a business who's in the business of selling that good or service.

43So you can keep diagnosing as narrowly and as granular as you need to for the user. And then the second thing after we diagnose is provide information. There are a lot of organizations out there providing information to the public to try and get them to help to try and get them the help they need to use the system. And here in BC, we call them Public Legal Education and information organizations, PLE organizations, and they do amazing work and they have a tough job and not a lot of resources and they manage to, again, just do great things. But what we try and do that's a little differently in using the Solution Explorer is we don't, we try to avoid having to send people out to different organizations and different websites and different offices and different front counters.

44We try and pull all the information that the user needs right into the system and not overwhelm them. So if you go to a public legal education information website and there's a 20-page PDF and it's only, you know, the third paragraph on page 17 that applies to your situation, the chances are you might not find that if you're somebody who's not familiar with the law. With the system, we can pinpoint that sort of information and deliver it. Directly to the user once we've diagnosed their dispute. We know this is the information that they do need, and we don't have to tell them about all that other stuff that was, that was around there. So diagnose, inform, and then we try and provide some self-help.

45So imagine, you know, asking a question like, are there things this user could do to resolve this dispute on their own? And if not, are there things they can do to start to manage it or contain it so that it doesn't get any worse. Some of the self-help is incredibly basic. It could be a communication template like a letter or email template to send to somebody, but we can populate— because we diagnosed the user's situation, we can populate that template with the exact information that a human expert has told us should be in this communication. The details about the consumer product, the date it was purchased, you know Whatever the human expert tells us we need, we can build that right into the system.

46So after the diagnosis, the information, the self-help, the last sort of bucket that I describe is streaming and triage. I kind of put them together. So streaming just basically means the system is set up to send the person to the place where they should go next to continue to receive the help from, from the system for their problem. In a lot of cases, the way the CRT's Solution Explorer works, that passes you through to the CRT intake process. So while you're using the system, you haven't paid a fee, you haven't filed a dispute or anything like that, you haven't started a proceeding, but you can be passed to the next phase if the interaction you've had with the system suggests that's the best thing for you to do next.

47There might be cases, though, where that's not what our user needs. And they might have something more urgent. This is where triage comes in. So if the user has a loans and debt situation and they've interacted with the system and maybe they've told the system that, yeah, I am never going to be able to repay this loan, like, based on my finances, it's just not going to happen. We don't want to send that person into a legal proceeding because they need to go somewhere else. They might need to go to a debt consolidation service. They might need to go to a trustee in bankruptcy. To talk about other things that they could do. So that's just one example, but the, the broader sort of description again is diagnose, inform, self-help, and then streaming or triage.

48And we're trying to create a system that can do a lot of things, like from a user's perspective, make it so that as soon as they start, they're in the right place. We take the, the burden off of them to have to decide where to go next. Or what to do, or am I doing this right? And when we try and build all of that right into the system, but from the Administration of Justice perspective, we're also trying to make it so that the people who come through are ready to receive that service. They kind of know what's going on. We've put some structure around their dispute and help them see, you know, what's involved, the way they need to frame things, and maybe even manage their expectations so that with this expectation management, some people are going to say, you know what, this isn't worth it to me to continue in this process.

49Now that I see how this all goes and what law applies, I might not want to continue. So I hope you or other people listening to this might be, especially who deliver legal services to people, might be saying, Yeah, that's kind of what I do. I give people some information based on my experience and my knowledge and help set their expectations about what comes next. That's just one example of a way that we can use a very basic system, and it's 1980s-era artificial intelligence, to provide some of this value to the public.

51TurboTax is one that people— or QuickTax— often people say is an example of an expert system that takes a complicated area and puts it, you know, makes it easy for the person.

54Yeah, I was lucky to be part of the team that came up with the concept and designed it and did the user testing and deployed it. It's— I'm not involved in the administration of the system now, but yeah.

56It's— yeah, there are terms of service. I won't go into them, but you can absolutely go ahead and read them. They just go through the system as if you were a user who is preparing to use it, and you'll see them there. If you're worried about, you know, liability or giving legal advice, it's something the team pay very close attention to, took it very seriously. But

58at the same time, we're trying not to replicate this thing that happens over and over again where public legal information will say Okay, looks like you got this type of dispute. This is complicated, you should go see a lawyer. We didn't want to replicate that dynamic. There are certainly tons and tons of places in the Solution Explorer where people can say, yeah, I want to get professional help, and then it will take you to a directory of where you can go ahead and try and get a lawyer. But that's not the only thing that happens. And we tried to share some of those sort of the knowledge that was appropriate with the system.

60Yeah, I could tie it to your other question, like the question that you asked, sorry, just before I went off on a tangent there about why it's, you know, maybe why it's not happening more. And it's culture, right? We're a profession and a field that doesn't take change too easily. It's sort of kicking and screaming. It changes. That's the way they happen sometimes. And this has been the story of online dispute resolution. This has been the story of using technologies to go in very, very cautiously, which is appropriate because it's new. But I don't think that an expert system needs to be limited to small claims. And again, as you note, the CRT handles personal injury up to a pretty high dollar amount.

61It handles condo cases with really no dollar limit. So it can handle big things, but it all comes down to design. And I, you know, you could design, you can design an expert system to deal with medicine, where people's lives are on the line. It's just a matter of inputting the right information. You could design an expert system to give information to a pilot on a commercial aircraft. The expert systems aren't unique to law, they can be really used in any domain. And so it's up to us as the knowledge engineers and the system administrators to ask the big questions. Is this appropriate? How do we design it? How do we design it so that it's safe?

62And what are some of the problems that we foresee and how can we mitigate them? I'll be the first to say that these, the systems aren't perfect. I won't make that claim. But the human-delivered model isn't perfect either. And so I wouldn't make that claim either. And there are different ways whether it's a human-delivered sort of service or an expert system or technology-based service, there are different ways that we can identify and address or mitigate some of the risks. So it would just depend on the issue, and I would say, okay, how do we design around this? Let's see if we can— if we can make it safer. So yeah, people ask me from time to time, just generally, about online dispute resolution.

63I think they can handle— I think they can handle the biggest, most complex cases. In fact, I think they'd be a lot more appropriate for some of the bigger, more complex cases. But again, I don't want to be in a dichotomy here. Part of the— part of the reason I say that is because I wouldn't just use technology. I would use a mix of technology and, and and more legacy-based processes, all within the same sort of matter or system. Once you mix these two approaches, or multiple approaches, things become a lot easier and safer and flexible.

66Yeah,

68Yeah, again, there's lots of different ways to— if we were dealing specifically with expert systems, there's lots of different ways to address or make— to allow for variability. So you can take someone using the system pretty far, but not go too far. And you can say, you know, this is just using my language. You would say it's common practice for this type of issue for people to do X or to do Y, but that may not apply in your case. I mean, there's a lot, there's lots of ways to, to do that. The bigger question though, that, that I thought was really interesting that you're bringing up, you know, drawing comparisons between medicine and law and why one has changed more than the other.

69I honestly don't know the answer, but I do find the question really fascinating. And I like to think about it. The one thing you and I could do as a thought experiment though, would be to take somebody, I don't know if we took some, somebody in our justice system who really calls the shots, you know, a very, very senior justice or court of appeal justice, a chief justice, and tell them, you know, we think we— you're going to need some brain surgery. And it's not a super risky operation, but they are going to have to open you up and go in with the scalpels and start doing some, some things to the— your brain.

70As you roll them toward the, down the hallway to make this happen, imagine that there's two operating rooms there. And Door A is the operating room that is framed on the way brain surgery was done 200 or 300 years ago. It still has the same tools that were used 200, 200 or 300 years ago. It still has the same practices. Maybe we weren't really into sanitizing our instruments and all of those sorts of things. You had the person with the big saw that didn't get cleaned. And or you can go into room B. We can wheel you into that room. And that room has brain imaging where we can see the inside of your brain and the problem before we even open you up.

71And we have these, you know, tools that use technology and magnification and monitoring, and we can control your anesthetic in a much more refined way even. What do you think that that very senior judge or justice is going to choose? And I hope, I hope we all know the answer. So just as an aside, like when the pandemic started, I remember thinking, wow, I'm sure glad that medicine embraces technology and change. And look what's happened, right? We have— they've developed vaccinations or vaccines in record time, I'm told. So, but law is different. Now, I think if I had to argue, if you force me to argue for why we won't change I think you're getting into the— you got into some of the interesting things about like, you know, we want a constitution for a state, we want it to hold up.

72When you have a bad actor who comes along and maybe gets elected to the highest office of that state, and starts doing things that are outside the norm, you want to be able to point to your laws and your constitutions and say, look, this is the baseline, this thing has to stay, you can't just go in there and change it and wipe it out and do whatever you feel like and recreate the world. So there are absolutely, I think, times and aspects of our system that we want to be more solid and immutable and want to hold on to that. But we don't need everything to stay that way. There are many different things we can do.

73We can do things differently, make improvements, but still hold on to those core principles. We can still hold on to that that foundation of law and the things that are really important. But we do, I think, need to open our eyes to different ways of, of getting there. And there are countless examples all across the system. So but yeah, hold on to that thought experiment. Next time you go into the hospital or need some medicine or dentistry, maybe that's a visceral sort of way to think about it. Do you want the technique that was done 200 or 300 years ago? Do you want to go see the dentist who refuses to consider something new because it might, it might not work, or it might be different?

74I think the choice is pretty easy there.

77When this comes up a lot, it's again a fantastic illustration of why we might not want to use technology. It might impair our methods for doing things like finding fact, or again, if you got two different witnesses, telling different versions of what happened, you need to— we have ways in the legal system that we've developed to decide which evidence you can rely on, which one to prefer. And this is where I really— it's such a great example. I would say, do we know if those methods work Like, it's one thing to say, hey, this is the way we've done it for the last 2 or 3 centuries, and we're not just going to throw that out the window for no good reason.

78But my— the way I'd come at it is what are we trying to do? Well, we're trying to figure out who's telling the truth. And then I would say, like, well, are there ways to measure it? Is there a way to use data? Is there a way to use science to see whether this approach works? I'd encourage you or other people to listen to a conversation that I recorded with an expert in this area from— he's from Canada, Dr. Brent Snook, and his— one of the areas where he has done some research and writing on is the ability of humans to detect lies in people or to detect deception. Is the way the psychology points it out.

79Turns out we're terrible. Humans are terrible. We— I think we hit about 53 or 54% accuracy when it comes to detecting deception. But we don't— that's a coin flip, right? But the scary part is we don't know we're bad at it. We think we're good at it. There's no way to discover that you're bad until you measure it. And that's what psychologists have done. They've done meta— analyses where they've studied subjects that included over 20,000 different subjects, and the data keeps showing we hit whatever, 53%, we hit a coin flip. What they've been able to do is find that people who are professionals when it comes to detecting lies— so you're, imagine the person who works at the border, customs officer imagine a judge, that sort of thing— they They're no better, according to these studies, as I understand them, at detecting lies.

80They think they're better, which might even make them a little more dangerous. So it's such a great example and reason why we, as a bunch of smart people working in the law, dedicated people working in the law, people who want the system to work properly, who want the system to do the right thing, who want the system to be able to figure out who's telling the truth and who isn't, why we should be open to new ways of doing things. For me, this would be talking to the psychologists. This would be bringing in a lot more science. This would be trying to find out ways that we can augment the human approaches to detecting lies.

81Imagine if we still let a human decide important questions like that. A judge, it's their job to decide who's telling the truth how much weight to give the evidence. But imagine if we could give that judge a lot more tools to augment her approach to the subject. So imagine if we could do a text analysis using an algorithm that has been, you know, tested a trillion times and seems to be pretty good at detecting lies based on the way people structure their language. This is real research that happens in some places. Or imagine if, I don't know, this gets a little bit scary, but you're hooking up electrodes to a person's brain and you and you know which parts light up inside the brain when they're telling lies.

82We could say we don't want to do that. There's something about that that doesn't feel right and makes our skin crawl. But we have to be honest with ourselves too and open, saying, yeah, we're going to keep on using the way— we're going to prefer the way that basically is the equivalent to flipping a coin. Do we want to settle for that? We can just— I think we should at least decide. Whether or not we're going to, to settle for that. We should have the best evidence in our hands, the best evidence available to us, before we say, yeah, we're good with the coin flip. And again, that, that example can be used over and over, lots of different methods, lots of different points throughout the system, not just for finding fact.

83But for me, this is where my view of the world, my view of the law is we've just got too much knowledge and drive and dedication to settle for those sorts of things. That's, I think, why that partner at my law firm— what he was, what he was putting his finger on is that we shouldn't settle, and we don't have to settle. I think we can, we can get out there and find ways to improve the system and make it work better, make it do what it's supposed to.

85Thanks, I enjoyed talking to you.