 And we're live. Hi, I'm Dazza Greenwood, a scientist at the MIT Media Lab. And this is one of the flipped classroom lectures for our upcoming computational law course. And I'm joined by the teacher's assistant for the course, Mila, coming to us from Brazil. Mila, why don't you introduce yourself? Hi, nice to meet you, everyone. I'm Camila from Sao Paulo, Brazil. And I'm very, very happy to be here, because I've seen Brian speak in prior occasions. So I'm really looking forward to it. Here, here. Great. And speaking of Brian, the lecture today is being offered by co-instructor for the course, Brian Wilson, who's chair of the Legal Hackers Chapter in Kansas City, Missouri, among other accolades. And he's going to be talking about providing an overview of computational law with a deeper dive into one of the excellent open source enabling technology, stock assemble. So with that, Brian, maybe we'll introduce yourself and get us into the topic. Sure. Thanks for the introduction, Daza. And thanks for the kind words, Mila. We actually got to speak in London earlier or later last year, and that was a lot of fun. But yeah, my name is Brian Wilson. I'm from Kansas City, Missouri. I work at a startup called Risk Genius, where on the operations side of this artificial intelligence outfit where we break down policy language from insurance policies and quantify it using various types of computer science. But what I'm going to be speaking to you today about is this overview of computational law. And with that, I'm going to pop over to the slides and go into present mode. And so what I really want to do today is just provide a basic and skeletal overview of what computational law is, how it can be used to solve for different legal frameworks or legal issues, legal use cases, whatever legal problem that you have, or at least improve it a little bit. And then we're going to walk through DACA symbol as a tool to do that. So what is computational law? A lot of people might not be familiar with it. It's kind of like a nebulous concept. Michael Guinnessarith from Codex came up with this definition. It's computational law is that branch of legal informatics concerned with the mechanization of legal analysis, whether done by humans or machines. And I think that highlights something that's going to be important throughout this lecture and throughout the course, which is that a lot of these technologies are great. And the reason that they're great is because they help people use the law better or people navigate the law better. And so what does that actually look like? So anybody who's been to law school will be familiar with this kind of structure for analyzing legal issues. So there's an issue, you apply a rule, you provide analysis, and then you reach a conclusion. So what I wanted to do is kind of come up with what that looks like computationally. And so I put together this graphic that shows, you can take a bunch of different pieces of data, you can input them into some sort of standard form. Maybe that's using just like Google Forms, maybe that's using an expert system, maybe that's using an Excel spreadsheet, could be whatever. It just has to be in some kind of like standardized format that you can then provide some layer of computation on top of. And then once you've done that, what you wind up with is some sort of output that allows you to see what the law looks like as you computed it. And so what are these kind of like ideas and how are they connected? So data is information that can be used to make some sort of decision. In the legal industry, this data could be static data. So that could be like laws, regulations, or court filings. It could also be live. So if your city's got an open data portal, you could use that to kind of like in real time, monitor things. If you have a compliance tool, maybe you could be monitoring stuff like communications, live and in real time. And it's really those building blocks with which you get to the final product. So input connects those disparate pieces of information in that standard form. So this will include requiring a lot of specification, harmonizing those terms of art that are business, legal, or technical. And so this is where a lot of the net, that kind of like legal analysis will come into play because you have to know how to structure it in such a way that it allows you to achieve the goals that you're looking for. Then it gets to kind of what I think is the fun part of it, where you're actually applying some level of computation. So this could be automatically sending emails, creating records, monitoring compliance, or it could be stuff that's more advanced. So maybe you want to apply machine learning to understand the variability of the wordings of your hundreds of thousands of contracts. Maybe it is using some degree of AI to come up with the optimal distribution of power in a, oh, what's that? What was Chris Barron's lecture last year? With the distributed power grid, it could be a lot of things. You could have a smart contract to record financial transactions on the blockchain. These are kind of where all the buzzwords come in. And then what you wind up with is an output. And so that output could be a dashboard, it could be a report, it could be an advice letter. These are where I think the most creativity will be involved. And then I threw this in as a bonus, but a lot of lawyers don't really know what GIT is and GIT's something that is very important to learn because it can help provide us with version control, like a system for version control that allows us to make sure that the files we are using are the right version that we want. It allows us to add context to documents and pieces of information. And that context is one of the ways that allows us to better compute these legal outcomes. So this is going to be critically important as things move into a more computational direction. And version control is important and because if you don't use it, you wind up doing stuff like this. I know when I was in law school, when I was just before I learned what GIT was, I didn't get it. I didn't get that, you know, there's a better way to do this. And instead of saving everything with final draft, final draft version two, final, final draft, you can actually save it with context. So the DC code, so the Washington DC is like series of laws, they had a typo in one of their regulations. And you'll notice it's this, I was supposed to be formatted as two Is. And what they were able to do was they were able to go in, propose a file change, say, hey, you know, this actually isn't the way, it isn't written the way that it's supposed to. Let's instead of, you know, keeping on, we can create a change for this, make a pull request, and then that file change is incorporated into the DC code for that section. And you're able to get the context of what the code looks like over time, why laws were changed over time, what circumstances led to the different changes, and you're able to have a more transparent process of understanding all this stuff. And so I'm gonna kind of walk through now some vignettes of what that framework can do in practical terms with these four examples. So the first one, TurboTax. TurboTax takes data about tax filings, or they have a lot of data about tax filings. A user inputs their own information. An expert system helps calculate tax liabilities. And then the output is that your state and federal tax returns are generated for you automatically. With Do Not Pay, Josh Bowder's app, they have data about student corporations, state and federal laws. A user again inputs their own information about the different services that they use. Maybe it's about companies that they've been involved with and then Do Not Pay in that computational layer, it figures out how much money they could possibly sue for. So you'll see it's over here, it's got kind of like a calculator. And then the output is that you can generate those claims automatically. So with impressive a button, boom, you're generating those claims. DLA Piper's got a great data protection map. That's really cool because it's interactive. And basically what it does is it takes data about the global regime of data protection laws. User can then choose to kind of like play around with, maybe I wanna see what the laws are for, what the laws are for that country. And then I wanna compare the laws of Canada with maybe the laws of the United States to see, okay, I've got this up. I wanna understand better how to structure my apps so that I'm in compliance with both of them. You can select the different countries, you can compare their laws. The user inputs like kind of like what problem that they're looking to solve for. And then DLA Piper has come up with a way to compute that. And then what you wind up seeing is this output of the requirements of those different laws and you're able to make more informed decisions. And I think it's important to highlight that this could be not just like a non-attorney, this is a great tool for attorneys because it allows them, instead of having to do something a thousand times, like generate world reports about what the data privacy laws are, they can work in kind of this object model sort of fashion where they're only updating those things periodically when they come out. And then they have an interactive way with which they can quickly navigate their clients' liabilities. And then finally, Relativity Trace is a really cool tool that takes data about insider trading and compliance regulations, monitors that data through inputs that are like a continuous stream of communication. So if you're a company and you've got a ton of emails, all those emails can go through this compliance tool and you can monitor that suspicious activity. Maybe there's some insider trading going on. Maybe there's, maybe you know, you're in a government and there's corruption. Maybe you're trying to monitor bribery. This tool takes some natural language processing, figures out what clusters of words might trigger some sort of notification that suspicious activity is arising and it automatically computes that in something like a dashboard or it can automatically send notifications to people. And so I think that's like a really powerful tool because it changes the way that people are thinking about legal. It moves from something that's static and something that's just kind of like this monolithic burdensome sort of like outdated tool and modernizes it in a way that makes it dynamic and exciting and accessible to people and does so in a way that increases transparency. So now people can understand what's actually happening. They can see the law instead of kind of living in the shadow of, you know, the billable hour. And that's pretty exciting. And so now we're gonna kind of get into what DACA symbol is, how it can be used as a computational tool and kind of wrap up by looking at, you know, some possible ways to get involved with using it. So DACA symbol is a program that, it's an open source kind of program that was developed by Jonathan Pyle. It's free, 100% free. It's open source. It's available to everybody and it is a way to do interviews and it's really approachable. I think that's the thing that will be appealing to a lot of lawyers. It's also really easy to use from a programmatic standpoint. So, I mean, all the computer scientists that I've shown this to, all the developers that I've shown this to, they're all impressed with how easy to use it is. And I think the key takeaway for lawyers might be that you can use this to automatically generate those documents that you have that you routinely file, that you routinely produce. And now you can just type in a few answers, click a few buttons and boom, it'll do that for you. And you don't have to worry about, you know, maybe mistyping some names or continually mistyping information. You just have it. And so DACA symbol can do a lot of cool things to compute the law. You can generate templates in DocX or PDF. You can use touchscreen signatures. You can create a chat bot, have live chat. You can process user input with machine learning. You can generate notifications with SMS email. You can even send faxes. I know that the legal industry hasn't moved as, it's not necessarily as agile as other industries, but this allows you to automatically send faxes. You can scan an image using OCR. You can create interviews in multiple languages. If you're involved in mediations, you can have some multi-party applications that different people use at different times. You can build whatever you want on top of it using Python. You can package your interviews in GitHub so that you can get feedback from the entire DACA symbol community. You can protect your client information, that sensitive information with various security features. So server-side encryption, two-factor authentication, document redaction, and it's also API friendly. So maybe you want it to communicate with your Google calendar or your Microsoft share drive. You can have it do that. And so you might be asking, how do I use such a cool technology? Well, lucky for you, I've got some steps. So all you have to do to deploy it, you can go to the DACA symbol website. And I've got the link to that stored in the slide. Basically all you have to do is, you can either set up a Docker container or you can use community.lawyer, which is a subscription service. I think it costs like $11 a month, which might be too much for you, but if you want to set it up using Docker, you can set it up remotely and it is free. And so I'm gonna go back to present. I've actually worked with one of the developers at the start-ups that I worked for and we created a Docker compose file so that all you have to do is download all of these different components. You run this command in the command line and it will generate your instance of DACA symbol for you. And we will have that in the GitHub repository so that everybody is free to use it. An important thing to note with it is that it only works for Max right now. We haven't set it up so that there are the commands for PC yet, but I imagine that's something that we will do. Once you've done that, there's a ton of cool functions that you can take advantage of. And those functions are the kind of neat tricks that you can use to generate whatever feature of your interview that you want to do. So you can take the user data and then you figure out the way that you're going to have that stuff input and this is how you figure out how that stuff is going to be input. You actually choose what goes into the interview and then you have the computational layer that says, boom, I'm gonna produce this cool feature. And so some of those features, I'm not gonna run through all of them, but if you need to send a fax, it will send your fax. If you need to send an SMS, it sends your SMS. If you need to format dates and time, you can do that. You can format numbers, you can do a lot of fun things. And so some examples, the example that they have on their website is for a custody complaint. And so this is just a screenshot of what you get at the end. You're able to download your complaint. You're able to download an advice letter. You can download it as RTF so that you can edit it. You can automatically have it email the right person. Another example that they've got on the website is a certificate of incorporation. So if you want to automate the business filings for your entity, you can set it up so that it automatically works with some standard form. You produce that standard form and then it automatically sends it to whoever you need to send it to. I don't think they have a way to mail it in the physical world, but anything on the virtual world, it'll mail it. So if you want to get involved with the DACA symbol community, there are a lot of great links here. They've got their own GitHub repository. They've got their own Slack channel. And on Community Lawyer, they have a DACA symbol toolkit that allows you to kind of operationalize this for yourself, or if you want some developers to work with who might help you use this tool at kind of like an enterprise level, they can help you do that too, which is really fun. But now we are going to get into kind of the, I'm gonna show you, once you get this up and running, you can go to the playground. And what's cool about the playground is that you can actually just start playing around with it. You'll see it's written in YAML, so all of the instructions are pretty straightforward. This is their Hello World demo. So when you run that, you see this page, you can click on buttons. Once you've clicked on the buttons and you're out of it, you can copy different pieces of code. So I'm gonna copy this code that plays around with guessing favorite numbers. And all you have to do is you delete that code, you copy it and paste it, you click save and run. And now we have this cool favorite number game. And so because I know what the answer is, I'm gonna click and guess what their favorite number is. And what's neat about the Docker file that we created is that it will automatically save this for you. So all you have to do is save this in your directory and then it automatically generates this file for you that is the exact same as the file in the playground. So if I just change this to what isn't your favorite number and we run it, now it's updated, I enter again the number and they say that's my favorite number. What's neat about that is it automatically updated that change and so you don't have to worry about saving or losing work. In any event, all of the stuff will be available in our repository. So if I go back to present, I think that is the end of the presentation. Kind of as we kind of get into more of a discussion, some of the things that I'm looking, I think we'd all be interested in getting feedback on and I think that will become more interesting as we lead into the third day of the course where we're actually playing around with this. We're kind of like wondering, what if we redesigned the legal system or just various legal functions using Docker symbol or like how might we re-imagine the way that the law was if we knew that it could do all of this cool stuff. And then finally, I think it's just really interesting. What are some examples of, like what are some of your favorite examples of things that can compute the law already? So maybe there's some apps that you use. Maybe there's some services that you use. What are those fun things that compute the law for you? And with that kind of, I think we can open it up to Dazza and Mila again. Thanks, Brian. That was extraordinary. And that definitely was something that needed to be said. So I'm glad that you were able to come forward and exercise some leadership to just give an overview with some great examples and then a go-to tool that's open source providing context and definition for computational law. It's definitely early days. You mentioned the Stanford work and Dan Katz has done excellent work out of Chicago, UMKC and MIT and other places have had a table. But we still don't have like a consensus widely agreed, standard definition or a scope of computational law. And I do believe you've just made one solid step toward achieving that. So thank you. Yeah, thank you. And I think something that I've seen in, I think you had that lecture from that conference in China, but like the degree to have things that or the need to have things that are interoperable is so important because if you can standardize things and have them be interoperable and exchangeable, you can start playing around with different APIs and different services to do things that are actually really meaningful. And I think we're gonna get into some of those use cases a little bit more in the course, but especially with the self-produced data examples that need to have things that can talk to each other is going to only continue to increase. Here, here. So maybe by way of, I'm sorry, Mila, I can't tell if you're talking, but you're on mute. Yeah, so this is just a funny saying here in Brazil, we always go like, oh, law is black and white, black and white. But now with computational law, it is at least three-dimensional, when considering all these layers and interoperability. So law is not black and white anymore. Here, here, yeah. Well, we've definitely added some color. There's a spectra of gradations and a flexibility for what you can do, especially with those integrations of the input and the output and the introduction of the concept of an API and application programming interface to allow data to flow across heterogeneous systems across different non-connecting, non-coordinating systems. So just picking up, Brian, by way of wrapping and looking forward on the last year of musings that you had on inviting people to speculate on ways that they compute the law now with apps or services, maybe a good question we can pose together to students that are watching this and getting ready for the course on January 15th when we launch it. So something they can think about would be what legal use cases can you imagine docassemble, as you just discovered it, could solve. And then in particular, a little twist on that might be for extra credit. And that could be achieved in a rapid prototype, maybe on that Thursday, January 17th, when we break into small groups and do team projects with the help of some MIT and other technical resources so that we could give life to ideas in short order. So what legal use cases, like legal situations, scenarios, vignettes to use Brian's words, could you imagine being able to do or solve if it's a challenge or with docassemble? And we'll go ahead and put this video as we're doing with all the flipped classroom videos onto our session page. We'll add a pigeonhole probably right down there. So if you look, or maybe it's above, I don't remember the user experiences. It's somewhere directly adjacent to the video. You'll have an opportunity to input your ideas on that and then don't forget to upvote or downvote or basically let us know which of those ideas you most want us to address. There's a little VODI button on the pigeonhole right next to the idea. So click your way to interactive contribution and that will help us formulate the priorities for the course. Okay, so before we close out, any anything else that you'd like to say? Mila, Brian? I think for me that that about covers it. The slides will be kind of a living document though. That might be an important thing to note. So there might be a few changes, minor changes I'm expecting, but we might add some more context, more examples, based on the feedback that we get. So get everything in that you want in and we'll review it and get it up there. You're here. And students, I think the golden rule is to interact. There's no right and wrong at this point. It's a matter of like proposing ideas and trying to hack together. You're here. So in that spirit, have a happy, hacky weekend and we'll see you online.