Law, AI, and the Future of Legal Practice with Professor Abdi Aidid

Hello everybody, Welcome to another episode of Studying Law around the World. Today I have the opportunity to be interviewing Professor Abdi Aid and we'll be talking about a lot of exciting things. Really wanted to thank you so much for your time and for the opportunity to be talking with you today. Professor Oh, it's a. Great pleasure. Thank you so much for inviting me. You know, you were so curious about everything and hopefully I can sustain your interest because I don't know if my story is anything less than boring. Surely it's not, it's, it's just very interesting to kind of like take a look at your bio and, and learn a little bit about all the work you've done so far. Practicing New York and Toronto, leading research at Blue Jay and and also teaching at UFT and now as a, as a visiting professor, if I understand it right, at Yale Law School. Can you tell us a little bit about your journey? Sure. Yeah. So I grew up in Toronto, where you now live, I guess, and I attended the University of Toronto for undergraduate. And, you know, going to university in the city you grew up in, sometimes you want a change of perspective, a change of pace. And so I actually went to law school in the United States and beginning in 2013 and I graduated from Yale Law School in 2016. From there I practiced law for a couple of years at a firm called Covington and Burling. It's really a DC, Washington, DC based law firm, but I was working primarily in New York for them. They have offices all over the world and I was working on litigation, arbitration, all manner of dispute resolution, mostly corporate, commercial types of disputes. I left then actually for personal reasons, my mother became ill and so I moved back to Toronto and I worked very briefly. I described it as having a very brief cup of coffee at a firm called Davies in Toronto, fairly well known firm that is predominantly corporate and transactional. But I worked in their very small and actually highly impressive litigation department. I mean, they were among some of the best lawyers that I encountered in my career. From there, I left and tried something very new, something entirely from left field, which was joining us legal technology company. Actually, I knew the founder and CEO Ben Allery earlier from many years before that when I was an applicant to UC law school and we met and he told me a little bit about this, this opportunity. And so I joined Blue Jay. And at the time, what Blue Jay was doing was using machine learning to try to actually predict case outcomes and was doing so successfully manner that was letting itself to use by lawyers, accountants, etcetera. So I joined to lead the legal research team at the time, which was a team of about four people, a lawyer and a couple of research analysts. And by the time I left some years later, it was a large team. It was the biggest team in the company. We had about 30 staff, including many lawyers, many research analysts. And that was a wonderful experience. Now, throughout that entire time, I was teaching part time at UFT. I was teaching legal research and writing, and then later a large lecture and Civil Procedure and a small course in arbitration. Around 2022, I joined the UFT faculty full time as a tenure track professor. And that's where you had, you know, we, I teach torts, among other subjects, Civil Procedure, arbitration, and I'm also teaching a course on privacy and data governance in the Graduate School. So it's been a very lucky and charmed career so far, and I take no credit for it. I will say, to the extent that I was involved at all, it was that I was open and curious for new opportunities and that that I think probably accounts for most of the the. Round that's amazing and and thinking about all of this rows that you've had especially related to technology and including the altering of a book on the legal singularity as well right I'd love to hear a little bit about you know thinking back on your time as a JD student and then having all of this incredible experiences what are some of the things that you felt that really shaped your perspective about what the future of the legal professional looks like and. That's a great question. So I will say this. I couldn't have imagined the possibility of working as the vice president of a legal technology company or teaching courses in law and AI, for example, while I was at JD student, right? In part because what you might call the proliferation of AI and in law really didn't begin at that stage. Even though I was in law school relatively recently, it didn't begin at that age. And so the idea of imagining a career in that space was a bridge too far for me. So I don't want to give myself too much credit. I wasn't, I didn't have that kind of foresight. Instead, I was someone who I think had very deep criticisms of the way that we do law and, and in the United States in particular. What do I mean by that? Well, think about the, let's call it the market misallocation, right? You have a high number of people that have unmet legal needs, right? So you have people who cannot afford a lawyer. And this is not just the problem that the poor suffer from, right? The poor have no hope of affording a lawyer. And maybe there's occasional government services that can assist them and voucher programs, depending on the jurisdiction they're in to help them get a, you know, legal aid, whatever, whatever it may be. But if you're a poor person, not only do you not have any hope of obtaining a paid lawyer, so to speak, but you probably have a limited choices. But lawyers are increasingly critical for navigating our institutions in our society. You know, it used to be the case that you maybe only needed a lawyer when you have a major dispute in your life, right? Or when you're charged criminally. Well, nowadays you might need lawyers to communicate with tax officials, you need lawyers to deal with landlord tenant issues. You increasingly need lawyers to do things like a state planning. If your, God forbid, you're in a situation where the custody of your children is at stake, then those that's not the kind of situation where you want to go at it alone. And those are still the big high stakes one, but there's so many garden variety social activities now that increasingly require a lawyer. For instance, people are using lawyers right now to reclassify their children into special education streams and schools because when they want their children to get extra assistance at school and get individualized education planning, they have to navigate a complicated procedural process, right? And so lawyers are critical now, the poor, the poor don't stand a chance, right? Then that can't be a just or equitable or ethical system right? Now, I started by saying it's not just a problem with the poor if you're a middle class, even if you're at the sort of a bottom end of what we might call the upper middle class, the cost and the expense associated with a lawyer can get so egregious, can get so egregious that it could bankrupt you. Certainly you're going to make choices that you you wouldn't otherwise make. It wasn't for the the that expense. And for me, there's just something that I can't wrap my head around, which is if law is critical as a guarantor of social rights, if it's important for us in terms of effectuating our interest in society, then why is it that it demands these, you know, very expensive intermediaries? And why are those intermediaries still expensive? So I was very curious for a long time about exactly what, what justifies the high expense associated with lawyers. And if you actually do some digging, right, you see that, well, part of what accounts for why lawyers cost so much is that lawyers offer specialized skills, right? They offer specialized skills that the market prices at X. Well, is it possible that given that these skills are helping people to navigate public institutions, to obtain benefits, to resolve issues in their life, is it possible that we can replicate some of those skills in a cost effective manner? And so I became very curious about the role technology could play. I just put a very fine, fine point on this. In my years as a lawyer, from when I started until today, you know, I'm still a lawyer, right? I was never able to afford myself. I currently could not, could not afford to pay myself the hourly rate that I would commend on the private marketplace. Something about that is concerning, right? So one question is, can we use technology to bring down some of those costs? So there's a couple ways you can do that. One way to do that is we can use technology to bring down the cost by making lures more efficient. If lawyers are more efficient, maybe they'll be able to pass on some of the cost savings to their clients. Or they'll be able to change the model from what you might call them a margin model where a lawyer makes money on a given case or matter they work on to a more volume model where what matters is that they have they're able to service a lot of clients and maybe they can bring the cost that way. That's one way. Or the other way is what you might call retail applications, allowing people to access technology directly to engage in some self help. So I became very curious about this actually really early. So I didn't quite know that I was going to be the midwife for these kinds of things, but I was thinking about it fairly early when I was seeing things like LegalZoom, you know, on occasion I would need to deal with my own situations and figure out ways to to, you know, self remedy, so to speak. And so that was the beginning and end of my curiosity with technology. And this is an important thing that I want to say. I'm not one of those people who's interested in technology because I like gadgets. You know, there's some people who are interested in AI because they aren't set in all technology. They're in the blockchain too, and they are interested in the latest release from Apple or Samsung, right? That's actually not me, right? I'm just a lawyer who's convinced that these things can be useful in limited context, especially for this access to justice problem that I'm talking about. And so I had a, let's call it a diagnosis when I was in law school about some of this. It doesn't seem to be working, right? But I didn't have the prescription until some years later when we started to see a I was showing us that we had some capability of reducing those. Costs, well, that's so incredible. It's it, it really sounds like you've had lots of moments where like the, the reality of what's going on in the day-to-day practice and then noticing what's actually needed. There was a lot of like, we need to do a lot differently. You know really this gap that you've noticed, I guessing so many, many areas, right? Yeah, well, let me say this, too. In law, we haven't done anything that any other profession or industry would recognize as innovation. So think about it this way. Let's just do a little thought experiment. Imagine a courtroom 100 years ago. It would be exactly the same as it looks today, except maybe today there'd be some screens for evidence presentation and some laptops on the council's table. Right. A law school classroom 100 years ago looked exact same as it does now. The class that you and I met in, for example, would have been structured the exact same way. And I, as a professor would have been standing up there largely saying the same things, right. And so there's an open question about that, which is, is it continuing to stay the way that it is because we have a bedrock belief that it's good enough or that staying the way that it is because we're not critically thinking about ways to make it better and new things that we can do. And so that's part of the the constant struggle of thinking about the improvement in the law is why are we doing it the way that we're doing? And you know, I have a hypothesis, which is that we've created a situation right now where actually hard to imagine other possibilities because law has what you might call a totalizing power, right? This is just to make it nerdy. If you have, if you have read Claudio, you've read them, you're gonna Habermas or something like that. You talks about the law colonizing the life world, right? The idea being that once law is introduced, everything becomes law like, right? Think about like a code of conduct at work. Once Upon a time, people would articulate a standard of behavior for you in the workplace and say, you know, Claudia, you better work hard and show up and do your best and do right by customers and be good to your colleagues, right? And there's ways that we can reinforce that behavior socially, communally, almost even like a family. Right now. It's the hypersexualized code of conduct, which reads like a statue. And there's an enforcement mechanism that attaches to the code of conduct. And there might be a grievance process and there might be appeal after that. And somewhere in the background is an Employment Standards Act that governs all of it, right? And so, so much of our life, you know, laws suffusing more and more of our modern institutions. And so now it's critically necessary to evaluate whether or not it's it's, we're doing things the right way, I think. And technology gives us that quote, UN quote crisis moment. Right, that's amazing. That's such a dive into in a little bit of the the history of it as well. But there is something that I think it's very interesting. So I actually got to read the legal singularity when it came out. I got you. I was here in Toronto. I got to see the AI tax man documentary followed a little bit of the events preceding it. And what I find interesting is that you mentioned a few things that did not change largely, but then there is the book itself brings a lot of things that did change. And it's especially how we classify information, how, you know, the law gets put together, how libraries look like back in the day, how much information you can have in a in a small spot now. But I guess the main thing itself, it's the same, which is the content. It just changed a lot of like how we deal with it, where we store it and and how we go about it, right? Yeah, so in the book we talk about these three eras of legal information, right, The analog era. The analog era is exactly what you know it to be. It's, you know, books, physical media and all of laws contained in those things, right? And, you know, at the University of Toronto, as recently as, I don't know, 20 years ago, we were training students to conduct their research principally by navigating the stacks of the libraries, right? And what's interesting about that is that it has significant equity implications, right? Because if law, if what exists, if what we know to be the law exists in physical media, then who has access to physical media makes a difference for the quality of the legal advice they can give. You know, there's a biography of Abraham Lincoln who said that Abraham Lincoln's legal practice was limited by how many law, how many books you could carry on his horse, right? Which is true. It's interesting. I mean, it's kind of a cartoonish way of making a profound point, which is that so long as it exists in the physical media that it's not so portable, then actually the quality, the entire universe of legal information is limited to that which you can touch and see. And of course, you know, same thing as today. You know, the well paid corporate law firm is going to have a better access to the information than the county lawyer or the local lawyer or the public interest lawyer or even the government lawyer, right. So, but beginning around the 1970s, you have this transition out of analog into digital, you know, the rise of the computer terminal, but also companies like, you know, W now Thompson, Reuters or Mead, who created LexisNexis, right? You get cool projects that come out of universities initially, like the Queens University information computing project, I think it's called Q UIC, now known as quick law, for instance, right? You, you start to see law become available and retrievable. These digital databases, part of what we argue in the, you know, you use them today can lead Lexus, Westlaw, Bloomberg, whatever they may be, right? Part of what we explained in the book is that that shift from analog to digital, even though it's a it's a shift from one era to another, It's not that profound shit. It's not the quantum leap, people say, because all it really happened is the information that exists in the analog area became platform digitally, but it was the same stuff we made the books and whatever cases statutes available digitally, right? Today we're in what you might call the computational error, where we say, let's take all the information that was generated in the analog era platform to the digital era, and let's treat it as data. And now let's perform functions on top of it, right? Let's train algorithms to retrieve to identify patterns in the information. Let's use now large language models to generate yet new information from what it's learned, for instance, in about the the semantics, the syntax addiction, the structure of that information. Let's do statistical analysis. Let's use advanced data science techniques to create new information, right? And so that's kind of where we are. That's fascinating. I love you mentioned the book and we're proud of the book, but we submitted the manuscript, I think like November 2022 and what happened that month, you know, change GBT hit the marketplace. Now we're observers. So we, we knew about large language models, generative AI, we anticipated that these things were happening. But it just tells you about sort of the rate of innovation here. Or when, when we first started seeing AI image generation tools in 2022-2023, you know, they were putting noses on people's foreheads. Today they could, they could trick you, you could, you could think an AI generated image is a real person, right? And so we've come that far and there are certain ways in which we're going to continue to improve. But I think we know enough now to figure out that we need to make some rules and figure out ways to absorb it culturally and socially and become better educated about it. All amazing. Thank you so much for your comments. Really appreciated that. And something that I that I wanted to ask you a little more about got you listened to a few of the different interviews you've given over the years. And something that you mentioned that I thought was just so interesting. Is that an observation or maybe a personal feeling that you have is really that AI will allow us to have more time to do a more creative lawyering. So I think some of the arguments you put out there was, you know, there's a lot of a lot of the, the work that we do, the manual work, so to say, ends up, you know, taking a lot of time. And then we, we don't go to law school thinking about, you know, writing a factum or, or doing these things. Might we might think about writing a fact to them, but we what we don't want to do is spend hours and hours in the Westlaw universe, right? Trying to find the one case that stands for the one proposition that the partner, whoever you're working for told you exist, but you're now suspecting doesn't exist. Right. Like these are the the drudgery involved in the legal practice are not things that we want to continue to do. Or if you want to continue to do them, you should feel free to do them, but at least with the understanding that there's more efficient uses of your time. And for me, this is a critical social imperative. It's not just about us having more fulfilling jobs of lives, though I think that could be a byproduct. I just think that we need to do things better, faster, more cheaply and more effectively, more accurately, because there are people that have unmet legal needs, right? Emphasis on the more cheaply, right? And technology allows us to get closer to that. Now it brings with it other problems too, right? I think the key thing in our book, the legal Singularity, which was an optimal, I should say optimistic, it was, it definitely advocated for the increases, increased use of AI in legal systems, right? But the subtitle of the book is very important. It says how artificial intelligence can make the law radically better. And the keyword is can because it's not automatic that it will. Why? Because there are still humans involved. We can still do things. We can impart our biases, we can discriminate. We can let this technology and I become a runaway train. We can fail to adequately regulate it. Or, and this goes back to your question, we can use it in ways that don't reduce costs, but but allow us to extract higher prices, right. So, for instance, you know, we live in capitalist systems. There's there's a there are winners and losers. There's also people that guard information to their advantage. If you're a large law firm and you've had a strong market position for a long time and you know that data is king in a world of AI, well, you're going to create your own internal white labeled AI tool that's going to be trained on your data because you have privileged information, You're going to have an advantage, especially in areas for which things like case law are not published, like in arbitrations, for example, They're firm. That's the leading arbitration firm. Why would you put cases back into the public repository if you've been on the other side of most of them? Why not just train your own internal AI and then suddenly have an information advantage, right? So I'm not, I'm not naive here in that I think it's automatically egalitarian. We still have to make choice. We have to legislate. We have to care about each other as a, as a critical bottom line, right? And so I think that's the sometimes, you know, I talk to people who are critics of the increases of AI because they are concerned about some of the potential for misuse. And I would say that's technology being layered on top. Existing problems that we have, and there's are ways in which it deepens and indeed amplifies those problems, right? Think about things like discrimination, right? If there's racial discrimination for a given immigration officer, right, but that immigration officers cases, case histories being used to train an algorithm that's going to then make immigration determinations, that's a problem. But it's not just that it's a problem. It's that suddenly the technology is more capable of discriminating at a higher clip and with considerably more volume than that immigration officer themselves were. So this amplification risk. So I'm not naive about this, but I will say, and this is the critical point here, we can't let technology take us off the hook either. I gave the example maybe in class sometime about imagine we had an AI tool that was setting our wages for the rest of our lives, right. So forget no more negotiating with your boss. He just you upload your resume to some tool and it sets your wages for the rest of your life, right. That would without any other interventions, it would project the gender wage gap into the future. Why? Because if it's based on all the historical data about wages, and it would find that women make $0.79 for every man's dollar, let's say, I don't know exact numbers, but let's say that now, if that, if that happened, we would, we'd be in the streets marching. We'd be like, you could not, you have to unplug this tech, right? Because if it, if it immediately spit out the this woman gets $0.79 and a man gets a dollar for the same qualifications, we would want to take a baseball bat to the computer. However, the gender wage gap persists socially. It's embedded in institutions. It's exist in a diffuse way. It's, it's, it's embedded in relationships. And there appears not to be the political will to completely, completely dismantle it, right? And so we can look at the AI and be like, Oh my God, we should never use it in that context. At the same time, we can't take ourselves off the hook for from not paying women more. And so there's a way in which technology is both a masking agent of some structural inequality, but it's also in some ways a good, it allows us to bring the issues into stark relief, just almost like an empirical tool. It's it's, it shines some light on some of the contradictions that we have in our society. So that's amazing. That's a, I really appreciate just you taking the time to have this conversation because it's been so many different interesting topics that you've brought up. And if, if I were to ask you then, you know, we, we, we talked about, I guess the practice of law, the information gathering, the social interactions that it had that, that, that it goes along with. And I know that some of the courses you teach are, are very much related to, to law and international artificial intelligence. So you know, when your students will come up and say things like, I want to work with the law and AI, what do you tell them in, in, in that prospect, If I may add to that question, I've been, I spoke with quite a few people that work with privacy law. And it's been more and more popular since the European Union's regulation of, of privacy and all of that. But before that, we didn't really have privacy lawyers, or at least they weren't called privacy lawyers. They were just lawyers that also did a little bit of privacy and this and that. And it just kind of like booms, right. As far as careers, will, you know, a way that you know, people were able to from my experience in these chats that I had, people who were new calls were able to take positions that beforehand in a previous market, you'd have to be a lawyer who is, you know, working 10 years out in the market. But because they're they're just wasn't a 10 year privacy lawyer, then younger attorneys were able to to get those positions. So. Is there some kind of opportunity like that with AI that? OK, so here's the thing. I actually think that there's a really good time to become a lawyer, right? In part because of what we talked about earlier, which is more and more of the grunt work is being automated. And so this frees up some time to do interesting creative things. And let's be serious, we came to law school because we wanted to do going to problem solve, be creative. We wanted to have a rhetorical flourish in the courtroom where we wanted to deliver justice. We wanted to help people, right, change the world, so to speak, right. So, you know, I, I think more of your dynamism will be on demand in your profession, which is really nice. The other piece of it is there's all these new things now. There's all these things that we couldn't exactly imagine. Same way I told you when I was in law school, I couldn't imagine being the VA legal technology company. Claudio, the job that you have in 6-7 years might not be one that's invented yet. You know, you might be a leading technology lawyer, but the technology that you're dealing with might be some newfangled thing that we can't currently contemplate. And so it's exciting just to wait and see in that way. But separate from that, just the technology that exists today is putting major strains on our legal system, right? What is what is what does intellectual property right now? What does privacy at the moment? What does torts if every autonomous, you know, every Tesla you see on the road that has full self driving is a challenge to our traditional negligence regimes, right. And so there's just going to be, you know, sad to say, a lot of legal issues generated in the short term because of these technology that that actually create, you know, to the extent they have a positive byproduct to create cool opportunities for you and me to to think about ways that we can use our skills. That combined with the fact that you can use the technology actually in your practice to be more effective and more efficient means that you can get good faster. And because, you know, you're entering A profession where people have, you know, we talked, we started this by saying nothing has ever changed your profession. That's pretty conservative and in some ways baked into it. It's mold and is unchanging unless it's forced to change. You can hit the ground running and have a bit of an advantage by having some technological, technological fluency, being able to leverage the technology in ways that are responsible, but also being abreast of the issues that it generates, Right. So I think there's great opportunity there. You know, there's going to be massive legal changes in Canada. You know, Bill C27, which was going to revamp our privacy law and also introduce the AI and Data Act, actually died in committee because of the election, because of because the government fell. But it'll be back and that'll be a wholesale change in the privacy law. And we used to have, you know, PIPEDA, we still do, but it's it's on its way out, PIPEDA and the Privacy Act. And, you know, there are lawyers who made their career from having been on the ground floor for that change a generation ago. And so if you're someone who's interested in being useful to your clients on the modern contours of privacy law, then you should observe what's happening right now in the Parliament with Bill C27 and learn about it even before it makes it to getting royal assent, because you will be ahead of the game. You'll be able to provide important advice. And by the way, any company that's worth their salt, that is serious about privacy law, they're adopting, they're adapting their practices to the new laws even before they pass. Right. And so you have an opportunity to say, you know what, I have this advantage. Like I'm someone who's inclined technologically, there's substantive legal changes that are occurring and I'm on the ground floor for it. And so you you can, you can actually, you know, make a life well, you know, from just they've been early in that way. And so I would encourage students who are listening, monitor developments, be up, be a relentless court watcher also, right? Because the rest of us in the profession, it takes a minute to catch up, just a minute to catch up. Yeah, fantastic. I really wanted to thank you so much for your time for bringing up this advice. I would. I would talk to you. I would sit here and talk to you anytime, even if we weren't recording. I really appreciate you, Professor, and thank you all for tuning into this episode and we'll see you on the next one.