 Hi, I'm Dad Greenwood, a scientist here at the MIT Media Lab, and I want to welcome everybody online to the inaugural day of the MIT Legal Forum. So we're just about to get started now with a keynote by Sandy Pentland. Stand by. I guess tell them to slide to me. Slide to me exactly. You like verbal? Everybody. Yeah. Can you just step? Yeah. All right. Hey, everybody. Well, I'm Dad Greenwood, and Michael and I, Michael, well, and you are, I'm physically, researchers here at the MIT Media Lab, who are, well, you guys are first aid. Very happy to welcome you here. Our research involves blockchain, law, computation, and this is our effort to reach out to the legal industry, the legal field, to help to catalyze the conversation around transformation of the field from the perspective of what I think we would call computational social science and the digital currency initiative. I'm not sure exactly how you'd say it. Michael, introduce yourself. Hi, I'm Michael Casey. I'm a senior advisor at the Digital Currency Initiative. I think computational social science is actually a great way to describe it because it is the blockchain that we're interested in. It's something far bigger than I believe than just the technological layer. It embraces governments, embraces governments and communities, and it folds that into software structure. So I think that's actually a great way to describe what this is all about and why we're exploring the interfacing of that with the law is going to be so critical in terms of how we actually get to implement some of the cool things that we need this technology to do. Okay, so mental note, you're also a computational social science. So it's spreading, Sandy. So computational social science is an application of analytics, data, I guess I would say in a sense, evidence in science-based descriptive and predictive models to social scientists to people. The field was largely created by MediaLabs on Sandy Petland and a number of other data scientists, social scientists, and obviously we can catch on. It turns out that the law is also considered a social science. I read it on Wikipedia. So when Sandy gives his keynote, a few moments, please listen for the advances, remarkable advances that have been made in other fields and breakthroughs with utilizing social physics, as we call it, to unlock potential of people in the economy and in education, transportation, so many different sectors and be listening. What could this mean in the fullness of time for the law? And then, following up Sandy's keynote, we're going to do some orientation. You'll hear some lightning talks of people that are describing breakout sessions and our methods are going to be to have people break into small discussion groups in the Bankwood Room around core areas of law and look at the application of blockchain and A.I. specifically and then the cross teams of identity. We're going to have a number of identity here who can guide us and contracts. So those are sort of our axes. And then see if we can unpack and reveal where some of the core, let's see, nexus points are that we can carry forward into your professions and into our research here and help to really spark that at another level of coherent, systemic transformation of law for the digital age. So with that, it's my real pleasure to introduce St. Helene and her boss, the top of her book, Social Physics. Okay. I just wanted to say a couple words to pick things off here. Let's see. Okay. So I wanted to sort of stretch your mind a little bit just to begin with before you get into more details. And what I want to talk about is this notion of social physics is where we got the notion of the census 200 years ago. The notion that somehow data and statistics, algorithms, could determine the way that we govern ourselves. And it fell out of use because it seemed a little too grand in 1840 when our statistics were just beginning. We didn't know much about surveys and so forth either. But today we have tremendous data. We have things to keep track of things like blockchain that are nominally interusable. We have machinery. So it's a whole new game. And that's what I hope to get out of here. You guys are thinking. And I just want to start off by talking about a couple of places where I'm particularly interested. I'm on the board of directors for the World Partnership for Sustainable Development Data. And what that unit is charged with doing is getting all the national statistical units of offices in the world to begin measuring far more of human life than they ever had before. Things like inequality, things like stress levels, et cetera, et cetera. And I'm also advising the American Bar Association on discussion areas. Can we set up an online learning program that covers all these new technologies and what should be in it and how should it be structured. So I'll be interested in hearing about that. The thing that drives a lot of this from my point of view is that suddenly we have data of everything, all this transactional data, but also all the things that come up, cell phones, credit cards, and so forth, which means you can know very fine grain, big thing about a person's life, about society. Yeah, there are people who aren't on a degree as they would, but that's rapidly changing and it's changing to a shocking degree everywhere in the world. So for instance, in the recent creation of the Sustainable Development Goals committee that I was part of put together in something that calls the World Accounts, it was pointing out that this data is something that we have to use if we're going to take on things like Sustainable Development Goals. Goals like justice, goals like equality, sustainability. We owe it to ourselves, we owe it to the world to begin to measure these. In fact, the Sustainable Development Goals include just shy of 200 measurements that every country needs to make. Every country is committed to make about the quality of their society. And those measurements use things like satellite data, like cell phone data, credit card data, like administrative data to be able to ask all of these questions. And what's really interesting is this is not just an informational resource, this is tied up to financial aid and to access to markets and other events. So large donors, many large international organizations are beginning to require that this data be done in an honest, transparent, and believable way before the citizens of this nation have access to international resources. In other words, international law is becoming data-fied. It's becoming driven by these new measurements. Well, you may say, okay, so that's a big thing with lots of discussions, but the same thing is happening everywhere. For me, having been in this sort of development area for a long time, I remember 15 years ago, everyone was talking about how half the people in the world had never made a phone call. And today, over 80% of the people own a phone, which means that you know where they are, who their friends are, you know a lot about their economic standards, a lot about their health, on and on and on. And you can begin measuring these things. So, for instance, we just completed a study with the government of Mexico about a program called FOSPERA. It's an aid program for pregnant women, but it uses phone data to be able to both measure and to advise, so it's sort of controlled, for pregnant women to be able to get adequate prenatal care and to make sure that their kids go to school. Something that we would endorse, I think, generally. I think if the kids go to school and so forth, they get a check every couple of weeks. But what this means is that these sorts of systems, which are automatic systems, we're going to have computers using data, steps to be able to get punishments and so forth, are beginning to spread everywhere. So, it's a continuous invisible nature of the three billion people's lives. So, in China, for instance, we just set up a new lab to investigate these sorts of systems. Many cities have no cash at all. Everything is on the phone. For all purchases, all transportation, all communication. And all of that data is used to determine your credit rating, your access, your social standing, your legal standing. As an automatic data... The other idea is used to determine your credit rating of life. Your access to social services is always in your program. Because, in fact, as an automatic data getting services to young kids and pregnant women, we require only to hear the notion that these same data are being used to ensure that people are quote-unquote a good citizen. The sorts of things that are happening. And of course, in this country, it's not quite as explicit as that, but through economic means, you see the same things. You begin to see more and more systems that are already added to that. And the question is, is how are we going to actually make sure that these are systems that are fair, have access to everybody, and what we want? There has to be some way of actually ensuring that the law keeps up with this auto-participated, driven evolution. So all interested in blockchain, all interested in AI, that's great. But those are the little building blocks of a much bigger system. A much bigger system is something you could call implicit or automatic law, where all parts of your life are regulated by measurements and views. But where do you think it's going to go? It's going to go, I can't go anywhere, it's going to open in as a presidency in the EU, work that we do in this area. And they're very concerned about this, in part because of the privacy law, which requires a great deal more tagging and tracking of data and much more explicit consent. And what that can lead to in the government law is a much more sort of big brother oversight. And the thing that was held in Estonia, and Estonia is a marvelous place in terms of digital identity and efficiencies out of the systems, but surprisingly, the way their system is set up, it looks like our version of 1984. The government knows everything, has everything, controls everything, which is odd for an ex-Soviet state as well as a religious natural for life. But anyway, so these are the sorts of issues that I hope that you'll keep in the back and the brain as we move forward. And thank you. I'll just see if I'm right. Very good. Great. So because any so concise, we have for a few more minutes until 9.30, let's take advantage of it. Some questions? Sure. Introduce yourself. I'm sorry. Let me say we're live-casting this onto our YouTube station. So be aware of that. And if you don't want to be caught in a stream or something, let me know. And we can actually create a place in the back of the room. But as part of our registration, one of the things that people were supposed to have clicked was sort of a video consent and a creative comments license. So if I won't put a room in the spot now, I'll be aware of that. And having said that, please share your question with the world. Thanks. My name is David Northpourst. I co-founded something called Innovation Jobs in search. Right. No problem. So my question is, I'm also a physicist. So I'm interested in the work of social physics. Now you said you have all the data and so on. What's the physics? Do you actually run real physics? So is it on the top of that? It looks a lot like spin glass now. So what you've got, except that it's heterogeneous. So it's not a uniform medium. People are not a uniform medium. And the other thing about it is because it's heterogeneous, the estimation problem is really different. The parameters for every item are different. But you're different than me, because which items can be estimated. But that's the type of thinking for it. But then what you ask is you ask about how do these things aggregate and you get an emerging phenomenon. So how do you get phase transitions and stuff like that? So the difference between, there's another technical term called sociophysics, which is microeconomics. And this is sort of microeconomics. I know that may have not been interesting to most of you, but basically it's possible to model a very fine-grained people's interactions. And as an example, something you won't probably know is that Amazon has a model of you and everybody who sells everything on Amazon running into that ground continuously to see what your typical behavior is and that's the way they interact for us. So they model in real time the entire world. That's not too much of an overstatement. It's the entire American world. So that they know where things need to be for shipping, what you're likely to buy with your credit rate. This is not so far from what you see in China, except they're using it for a benevolent purpose, which is detecting fraud. And in fact, they have almost an order of magnitude lower fraud than most of the payment companies because of this. And that's where the end-of-life product comes from. That's the end-of-life product. It's going to need to withstand those fraud by modeling the entire country. Okay. Okay, that's fine. Hi, I'm Kera. My name is Benjamin Pan. I'm from Davogator, Shanghai. My question is that you just mentioned you set up a lab or center to analyze, study the Chinese data which had was social media. I wonder, as a foreign institute, how can you do the research or studies over Chinese data information because it's really highly recreated. So the lab I set up is sponsored by the Secretary General and the former mayor of Beijing and sponsored by China Construction Bank, the former head of China Construction Bank. So it's their way of setting up ingesting knowledge of outputs and social visits and some of the reasoning and some of the systems that are being developed in the parts of the world. And from my point of view, I think that there's too little dialogue between China and the United States and it would be useful to know what different communities think about some of these same issues. Sorry, just to pull out the policy issue, I have another question that as we know cost loss of workforce and the money to generate the big data, now there's so many database available but some of them are just restricted by user commonly or government. My question is that is it possible for the data in future to be shared among all the colonies or organizations if it's not a privacy problem? So the systems that we are deploying which are sponsored by Princess France, the World Bank, people like that are for example. So in Senegal and Colombia by the end of the year, a data sharing system that allows private data from telcos and banks and other people to be made public, very much like census data so that you can see where the people are and where they go in aggregate. And that will be used for transportation planning or income inequality programs or things like that. The general structure is being sponsored by a variety of companies and governments because it permits greater data sharing. The key thing there is twofold. One is that data sharing, the principal thing that's difficult about data sharing are the legal questions. I'm not going to give you my data. That's not going to happen. But I might be willing to answer questions about my data. That's a much simpler legal negotiation. And so that's called a federated data system and that's what we're creating for some of the people here in the room. And what that allows is much more liquid sharing the data for whatever purposes. Also, I have done correctly it allows auditing of the data transfers which means that you can have human oversight to be able to judge whether algorithms are discriminatory or fair or what not. Because you can see the questions that were asked, the data that we've used and the answers that you've generated. The system is called Open Algorithms. You can see it online if you'd like. It's r-p-o-p-a-l-project.org or here at MIT the infrastructure is trust.mit.edu All right. Thank you, Sandy. All right, Sandy Peppin. One more time. And we are thanks to a Sandy's schedule conflict I think that we can be about 20 minutes ahead of schedule and my suggestion is when Mike and I talk you through the flow of the day we'd like to provide two things. And allocate about half of that so you can hear more of these lightning talks about the specific topics and breakout sessions and the facilitators that we have and actually give a little bit more time for Q&A so that we can make more powerful choices after lunch about what groups you'd like to sit with and also to maybe help surface the core ideas or problems or questions that you have. Honestly, so many people I want to talk to you out here. Hi everybody, by the way. I'm really glad to have you back up here. Awesome. Thanks, James. All right, so come here, Michael. All right. So, Keith, what we do here, where is this method of workshop that creates a program? Sure, sure, sure. The program is a great way for us to collectively draw out the sort of joint wisdom, if you like, in the room in the sense that it's kind of like a lazy form of academic work to get you guys to do all the thinking for us. But collectively, our rain power should be more constructive than just having one person talking at you in the room. So the workshop method is to sort of lay out a set of problems to talk through the parameters of what they are to then, in the process, you can explore some of those questions and we kind of effectively whack them and hopefully come up with ways to find what certain solutions to these problems can be. Of course, in this case we're talking about trying to put legal definitions on some of the emerging concepts that are coming out of this space. Trying to find ways to tackle from the legal perspective what some of these new concepts are. So to give you a sense I'm like involved in the AI side, I think I'm the blockchain guy. We are grappling with things like the notion of digital assets for the first time ever thanks to this concept of the ending double spending that we can know that we can create something of value that can't be replicated for the first time ever. And we have digital scarcity which means we have a digital asset. But how do we define that in legal terms? It's one of these issues. There's jurisdictional questions that we want to be exploring about where does the buck stop, where is the authority fly when you have a blockchain that is validating facts anywhere in the world, at any given time how do we grapple with the idea of where that data, where that power, where that authority lies. There's questions around property, there's questions around ownership that come from the notion of an asset being represented in digital form and then ownership with control of it being applied to the control of your private key. Where do we draw that distinction between the passage of physical possession and ownership with this very sort of digital dematerialized concept of what it's all about. There's a whole lot of big questions and they feed their way into some specific use cases which we'll be dealing with in the workshops. Thank you. I'm glad we're recording this. That was sort of best of the stuff I get all the time here when people look at me and they ask me like they want an answer to ownership. They want an answer to jurisdiction. They want the answer to the questions of even encapsulating the rights and obligations of parties in a system and automated entity that nonetheless is legal to pursue. So we've met a lot of ringers here that are on our team for the law. As you can, Kerry is with us now who's a member of Sandy's team as well. Former general counsel at Department of Commerce and does a lot of work in some of those issues. We've got other folks in the room. Gabe Tenenbaum who's a visiting professor from Suffolk Law this semester in Sandy's group. Jonathan Askin professor of law in Brooklyn Law School who's visiting professor last year. Many others. But we don't have the answer. So there's a little bit of a selfish motive as Michael was saying which is that we're doing a little ground breaking work here across the media lab in Sandy Pentland's group with computational everything and in digital currency initiative with blockchain and everything. What we would like to do is now have more of a regular liaison with the field of law to talk about in a structured way what some of these Michael called use cases we're going to call them scenarios today but they are and to talk about them in a way where we can all play. And so our method is going to be and I hope you can see up here what we're doing next and we'll accelerate a little bit is we'll set the terms of the day and then we're going to do some lightning talks some scenarios that are under discussion. Our method is thinker scenarios, everybody. So one scenario is data protection. Another one is legal entities. Bankruptcy supply chain from that energy immigration, litigation state crowdfunding. What these things have in common is you'll notice they're not they're not every one of these is basically a rudiment of law so you might see in a first year law school class property, contract bankruptcy legal entities. This is a good place we are speculating what I explore with you to ground a use case in a legal class pattern and some business context so that we can have a shared reference point to anchor discussion in this way we are seeking to be able to unleash everything to MIT people and other technologists and the legal people can bring and be talking apples to apples to one another. We start from a well settled well understood stable area of law with like two or three words and then we build the use case on that so Michael will talk soon about a very interesting solar grid blockchain use case where you could be thinking we might have said public utilities or we might have said a number of things we called it energy was the way that we came at it but we want to start to answer some of these questions in the context of fat patterns and the last thing we'll say is that because always with an answering question you always say it depends and depends on whatever it depends on that depends on what other things depend on as a matter of just scope to integrate legal and technical use cases for business cases we've been looking at bar exam questions and law school exam questions specifically bar exam questions and strip mining a bunch of them in each of the areas in some areas that we're not doing this today to figure out the level of abstraction and the kinds of questions that are answerable after the story like what are Alice's rights against Bob what are Bob's defenses can Bob get his car back can Alice get a refund we know that these kinds of questions are answerable sufficiently so that the field of law will educate that you can practice law after our exam I'm just going to go on a small limb to say that ought to be good enough for us to be able to say we have now incorporated all of the dependencies we know the main factors that answering this question ought to depend upon legally the main ones and that's to have a reasonable conversation about legal advice but to achieve and here comes so listen to the next one two words three words predictable legal results that's where we're heading so by the time we get to the end of Tuesday we've got the report back from discussion groups with these great technologies around it in common contexts the more we can share good ideas or even open key questions we need to answer in order to achieve predictable legal results the more I believe we will fulfill the potential of getting together today and having a conversation that is going to be worth continuing taking back to your various context and fields of law firms and professions and also perhaps continuing on in different ways we'll hear about tomorrow we'll say later today at MIT so that's the big level of what's happening so now so thank you Michael let's thank Michael again I'd like to ask Christian Smith to come up we're going to go a little faster and also from IBM Watson Brian and Shawna or actually you can come up when Christian is done so the next session is kind of two come on up Christian let's see if she slides up and I promised you that we'd be doing some level setting and so so we are let me get down here so this session is fundamentally about I think you're hearing here's the click so what we're trying to do now is first Christian introduce yourself but he's what might go to engineers here at MIT and in the media lab but a lot of rapid prototypes something with blockchain with authentication technologies especially and great guy and understanding stuff we were talking with Bob Craig from Baker Hostetler and also a member of a program committee who we will all meet for lunch about the program when Christian was talking about some common myths and some misconceptions or some confusion around what blockchain technology is what it does and doesn't do the sorts of capabilities it has that are well suited for business problems and even for legal requirements because we can do quite a few prototypes like that and the kind of capabilities that are well suited that just don't result and blah blah that was amazing that was great and so did other people as we socialized and so we decided to have a quick session between us starting with Christian here on blockchain and then Brian and Shauna from IBM on AI about methods for eliciting methods for distinguishing with technology does and a method to identify good fits for technology and and business contexts so let's start with some level setting of blockchain Christian here on it's working so there was a less polite title suggested by some people for this talk and this is what we settled on okay I told the purpose of this conference is to help transform law in a beneficial way and it's happening now because there's a technology that's particularly suited to many aspects of the law I'll qualify that and say I'm not a lawyer definitely a hacker so I don't know what I'm talking about there but there are some serious risks that come along with these technologies too and that's why I was asked to come here to highlight some of those things that need to be better understood and the people in this room as I understand it are critical part of the adult supervision that's helping to shape our future so it's actually important that everyone here can cut through the hyperbole and exercise the vision when it comes to aligning technology so we're going to talk about some pitfalls and misunderstandings in blockchain and in particular in blockchain and this isn't to discourage anyone but hopefully it empowers you to think better about this stuff going forward and given the tools everyone here is quite capable of distinguishing between signal and noise so let's take a sober look at what blockchain isn't isn't, what it can and can't do why you want to use it and how to evaluate some of the wild claims that people make about it to kick it off and question for you wait so what do you think the blockchain is someone asked me this about a year ago and really stuck in my head this is a terribly misguided question I don't need to pick on a person to ask it because we all start somewhere but it does illustrate incredibly well the degree of confusion we have in the current public discourse this question is struggling in two ways first, fuzzy terms like the blockchain are really vague and overloaded to the point of being meaningless and second, the definition of a blockchain should really be all that philosophical debatable or open to interpretation all this vagueness and subjectivity is a real issue it's hard to talk about concrete problems and opportunities when words can be just about anything so we need to define something to say blockchain is not a proper noun we talk about cryptocurrencies and smart contracts, digital autonomous organizations consensus protocols there are lots of variations like the idea of distributed ledgers and there is Bitcoin and Ethereum, Zcash lots of hyper ledger projects like Corda, Fabric and Indy but there is no such thing as the blockchain at best this is an overgeneralization at worst it's a false idol the blockchain is about as real as Santa Claus the Easter Bunny and the Tooth Fairy and there are probably some people that want to kill me for saying stuff like this but I'm saying it anyway it just doesn't exist in the way that so many people use in time these days so what the heck is it well first and foremost it's a data structure it's a data structure that can be cryptographically verified to ensure its integrity the cryptography used is not particularly novel we've had the main and derivatives used in blockchains for decades the most important ones are hashes and signatures hashes are used for verifying the integrity of data signatures are used for verifying the authenticity of it the data structure we build up represents information chronologically in the order it's added it serves as a tamper proof ledger we share a copy of that with everyone concerned it could be open to the world or it could be access controlled either way the ledger is replicated in its entirety between many parties in kept in sync over time the particularly novel thing about this is that each party with a copy need not trust any one of the others open public ledgers use a multi-party algorithm called a consensus protocol the consensus protocol prevents bad information from being recorded onto the ledger it ensures the validity, integrity, authenticity and consistency of the ledger across replicas over time with Bitcoin this protocol solves a specific problem called the double spend problem you don't want people to spend their money twice you want to dig deep on this read up on business team fault tolerance there's a working definition of blockchain let's reiterate we can talk about business data structures, cryptography consensus we can talk about legal requirements security goals and threat models we can talk about specific protocols and software and networks but let's not talk about the blockchain in the abstract there's no such mystical superpower that absolves mathematics and the laws of physics and solves every problem we feel like invoking it over so now we have a working definition what should we use it for some startling news ledgers don't replace every other technology that ever existed they're not suitable for solving every single problem and they're particularly poorly suited for solving some problems the overwhelming focus on distributed ledgers at this time in history is understandable because they are amazing but it's also leading to some absurd applications have you guys heard of the latest celebrity ICOs the New York Times should get understatement of the year where it's not always clear what they're selling it's probably fair to say right now if what you're doing in technology doesn't somehow involve blockchain it won't get any attention at all because of that everyone has kind of perverse incentives to blockchainify all their ideas and ventures and products and it kind of reminds you of the old saying when all you have is a hammer everything looks like a nail there are a couple simple questions that we need to ask about whether or not we should use a tool one is can we can we use this and sometimes you can but it's a not necessarily a good fit this is swine and flies with a sledgehammer and the next question is should we sometimes sometimes it's actually dangerous or harmful to use a tool for something that we may not expect or be able to predict all too often people ignore the bad fit that they identified with the first question and then skip the second one altogether we'll dive deeper on what you shouldn't do with blockchain first let's cover some important things to keep in mind when evaluating products for specific use cases every new technology goes through a hype cycle in the beginning nobody cares about it then something really hits in the come stuttering in everyone starts reinventing everything they do around the market gets ahead of itself and people start taking crazy business and financial risks Alan Green's band called this Directional Zuberance if you guys remember back in the dot com days and you'll know what the stock card looks like it's a big hockey stick it comes crashing down usually just when you start thinking this time it's really different then there's the bust so there are a lot of examples of this kind of herd behavior throughout history in 1637 there was tulip ania people got really excited about speculating on tulips there were gold rushes in the new world the same kind of thing happened there's the ai room in the 80s dot com bubble in the 90s we had a housing bubble in the last decade and we could be leading up to something like that here as well it doesn't mean that there isn't great we may be getting a little bit ahead of ourselves the next thing to think about is fear of uncertainty and doubt this particular hype cycle is scaring the crap out of some deep pocketed companies and industries even governments nobody wants to be the business of selling handmade buggy whips when upstart competitors are using assembly lines to crank out automobiles so people are wondering if this stuff is for real how does it work should we replace us should we buy in do we even have a future who moved our cheese there's a lot of fear and a lot of money behind it and people are looking for answers so it's kind of an existential crisis for many companies that are sort of vested interests and there's a kind of hustler that preys on this to the next topic which is snake oil when you put fear and hype together with an excess of money that could possibly go wrong type like these being a fog and mirror or you can get funded and so it's snake oil salesmen and earnest but overconfident people that show up along with the very authentic and genuine ventures so easy capital it empowers charlatans and dilatants as well as as well as the genuine article so you guys should pay attention and think about that not everyone is but at least consider it and the reason is because at the earliest stages of any project or venture it's very difficult to distinguish promising work from weaker attempts and that is because of favor not all marketing around undeveloped or underdeveloped ideas is bad it's valid to test the market for a product or service before you take big risks and we need to assess the level of interest among early customers before we can go and get capital there's an entire school of thought around this it's called the lean startup and this is totally legit it's actually a scientific method applied to creating new businesses but we have to recognize and many of the things we hear about aren't real yet and some never will be so brings us to another question how we know what's real and what's not when it comes to distributed ventures I can give you a couple hints obvious one is repeated references to the blockchain without any other meaningful context might be suspect circular arguments that answer why blockchain with because the blockchain blah blah blah if you look at those very carefully sometimes they're pretty sophisticated but they can be circular and another one is when someone says just hash it and put it on the blockchain this one is so common I'm devoting an entire slide to kind of a joke among hackers around here but it's also a good heuristic pay extra attention if someone says just hash it and put it on the blockchain recording a hash of some data by itself is completely useless the context and purpose matters and the asset test is to ask why if someone can't give you a solid reason and explain how this solves a specific problem they're all going to start in the photograph the most important thing with the noise though is to listen to your instincts you have the same intrinsic powers of reason like everyone else and they are good in fact most of you are probably developed them much farther than the average person so if you look into something and it doesn't make sense if you can't connect the dots if it doesn't feel right at a gut level there might be something wrong and it's not you it's them so trust your instincts um and I'm missing a couple of slides here I don't know how that happened because of the ranting for a minute good things is a ledger potentially good for proof and transfer of ownership so it turns out to be pretty good for money we know that one it's also good for things like title and land registration transferable records DCI folks are doing some really cool work on that talk to Mark Webber if you get a chance he's a really insightful guy securities are a good application another one that I learned up was diamonds so the provenance of diamonds is kind of a contentious issue and there's someone I think it's called Everledger that's doing something with cracking diamonds on a ledger and it's awesome another good thing to use a blockchain for is proving the order of events so it's really hard to guarantee exact times in the completely 100% approval sense but proof of sequence is another thing we can get very reasonably close to good times and we can definitely get the sequence of events pretty close as well so if you want to know who did something what they did and when they recorded that on a ledger and it's pretty good at that another one is logging for transparency so there are some really interesting cases for this like developing countries that have corruption problems companies that have data and people are using that data they're requesting the data and using it for not necessarily good purposes but when you log the access to that data you can kind of prevent some of that especially if it's public ledger and it's tamper proof I'll keep some of that in check I think Estonia has a blockchain called Guard Time that they're using for this purpose and as a citizen you can go and look and see which government agents I may slightly off on this but which ones have access to that this is a really powerful concept and this is good because having the immutable public records really helps to decrease corruption and fraud the bad let's go back the other direction if you google good uses of blockchain most of what you'll find is worse than useless the web is like a big echo chamber on this topic especially at a business level but even in technical circles there are things that have absolutely nothing to do with ledgers and smart contracts or being promoted to blockchain applications there are some things that ledgers are not good at like it's not a good communication bus you're not going to use it in place of signal or Slack or Skype or something like that generally not all that great at storage and retrieval they're optimized for protecting the integrity of the data not necessarily querying and scale is another problem eventually maybe we'll solve this one but right now if you think about credit card networks there are tens of thousands of transactions per second and Bitcoin is limited to 7 and if you change the parameters there you reduce the goodness of the consensus algorithm effectiveness so then there's this ugly part that you really need to talk about which oops which is confidentiality in blockchain a blockchain has no special powers to preserve privacy in fact most widely used ledgers are public they're designed for transparency and so that everyone can verify the correctness authenticity and integrity of the data they are not keeping secrets it takes a great deal of care to use them without somehow identifying yourself by correlation the instant you record any secret or personally identifying data on a public ledger you've compromised your privacy and potentially your security if there's anything you want to keep secret it almost certainly should not go on a public ledger encrypting the data that you want to put there doesn't really solve the problem either first it just creates a new very difficult key management problem when you encrypt something in order to decrypt it you or the person you want to share it with needs to have a corresponding decryption key in order to keep the encrypted data secret you need to keep the key secret if you want more than one party to be able to decrypt it you have two bad options you can share the key with multiple parties this is as insecure as sharing passwords it flat out defeats the purpose the other option is to re-encrypt the data with different keys for each recipient but this is also a bad idea because remember it's expensive to record things on blue blockchain on purpose it's part of how we keep the integrity of it so you actually don't really want to record the secret once level of many times the second reason encryption doesn't necessarily save you is that cryptographic algorithms tend to be predicated on computational infeasibility of reversing them by group force an algorithm that's considered safe today is unlikely to be safe in the future depending on how far you look down the road in a way encryption is a race against mortgages in addition to that there are always people trying to crack algorithms and sometimes they succeed if you want to keep a secret don't record it on an immutable public ledger I want to share with you guys the worst blockchain idea I've ever heard this is the worst one pay attention to this putting your decoded genome, your DNA on the blockchain because health privacy yeah do you think that's a hip-hop complaint maybe is the blockchain a covered entity what are the things that you should be looking at here what are the opportunities and the good that we can take from this if you look at legislation and enforcement of law very often we're dealing with designing disincentives for that behavior punishment one of the things most exciting about this new class of cryptographic protocols publicly under the umbrella of the blockchain sorry blockchain is that we can design incentives for good behavior while mitigating a broad class of bad behaviors they just can't happen we design systems that prevent them and then we reward people for doing the right things this may be a fundamental advance in the nature of how society is organized we can't change human nature and we can't design perfect systems but we do in order of magnitude better using distributed tamper proof ledgers so there's a desperate need here for critical thought I want to ask everyone as you're working on stuff in the conference and ideas and legislation and whatever else you guys do consider what could go wrong very carefully think about unintended consequences it's sometimes helpful to explicitly think about that and think a few steps ahead if you have this then what are the possibilities that open up from there so I'd also challenge everyone to learn the basics of cryptography and information security at a high level it's not all that hard they're just a handful of concepts and learn to think in terms of security goals and threat models that's also very helpful so I think that's that's it we really need law to align with technology now more than ever so please give it your best thanks not so fast Mr no no no he's shy blockchain demystified so myth busting, some level setting and hopefully some good ideas about how to recognize a business case maybe a legal fact pattern let's check for which blockchain capabilities are a good fit and maybe not such a good fit alright so first step right so we encapsulate blockchain and then next up will be great Brian and Shana and you're getting a lot there so let me see if there's any Q&A or questions, comments ideas let's meet halfway so name and quick question let's from Strickland Strickland in New York Carmel Wallace from Strickland Strickland in New York that's great the topic's great it seems to me one of the core issues is identity whether you have an encryption key, someone else has an encryption key whether you're party to an agreement can you talk a little more about where you see blockchain helping with identity yes so this is pretty interesting because from a read from a software that's here somewhere and those guys are doing amazing work on this, they got a grant from DHS that was about finding the applicability of blockchain 2 identity and before I came to MIT I was working on an identity startup and also was applying for that grant and I kind of personally gave up on it because we got into it and I didn't know how to apply for a grant and say it doesn't really apply except for maybe these two little things that it turns out to be really good for and Drummond got the grant and those two little things like proving key ownership or disseminating keys and also revocation so revocation of key, like if someone steals your key that's compromised you need a way to tell everyone and the best way to do this is to have some data structure that's replicated and you know it's verifiably good and everyone can have a copy of that so I think blockchains play a small but very critical role in identity and we need to be a little more cautious about what we try to do with identity on them you can't just take your driver's license and throw it on the blockchain and you know all the things that go on is it good? Anybody else? Okay one more Great Can you raise your hand again? Oh great Name and fairly rapid question Hi Judy Reiner What about using permission private blockchains for holding data because that's what I'm hearing about Sure That's certainly a valid kind of thing The most interesting part of blockchains is the fact that you can have them be open in public or not necessarily open in public but between untrusted parties and they don't trust each other then you need this kind of fall-tolerance algorithm but even in these cases where you want to keep stuff private it's important to think about data minimization What if you're hacked where do you want this data to live where do you want it to be a single point of truth and be the place that is protected from harm It turns out if you start to equibax if you compile the data in one place and it's all very very sensitive you create a honeypot people want to get at that they want to attack it so there's an incentive to try and break that and steal it So it's still useful and valid to think about what kind of data you're reporting on the ledgers and who has access to them even if they're private Last question You know the drill Hi You said that encryption is erased against Moore's law and I've also been reading about quantum computing and how it has the potential to crack encryption keys in seconds in the future when that becomes a viable technology So my question is how do we protect sensitive information in the future Yeah, so some of the concerns about quantum are overblown and some of them are really valid That's not my area so I don't really feel I don't want to put myself way too far out there on that one but there are algorithms that are known to be more less safe against quantum tech attacks In fact the bitcoin curve the elliptic curve bitcoin is the elliptic curve cryptography and there's a specific curve that it uses that has certain characteristics that are good for blockchain and good for the ledger but it's not necessarily the most safe curve and in fact it's suspected that it may be vulnerable to quantum photography and for that reason you'll notice that public keys don't appear on blockchain as much as it dresses when the dress is basically a truncated hash of a public key it's intended to not disclose the public key in a way that makes it vulnerable So we still have to mitigate those things We learned that the hard way on that people talk about it lunch tomorrow we did some rapid prototypes we did a digital signature prototype using the key pair from a classical public blockchain to say algorithm curve that you could coin that they're in sccp256k1 and we discovered the hard way that actually sharing the RSA is sharing the public key that's what I learned in the 90s at the American Bar Association Digital Signature Guidelines it turns out that's not always good advice but some curves doesn't make sense I'm glad that you asked that question we can start to surface some of the limits of some of these core technologies like they did in terms of the autonomy of the top of mind lawyers because by public keys I identified it it's also the way that you can encrypt things to people it's the way people sign things signature, legal good to know so let's see how many of these judges we can service with one another in our Google slides so that, thank you so much for spreading these slides out making cartoons and laying it on the line for a lot thank you Kristen Smith and since we're on a roll it's my extraordinary pleasure to introduce friends and colleagues who teach me a lot about artificial intelligence from IBM Watson and I'll be able to introduce these cells Brian and Shana but I do want to say quickly as people come up to our members of our program committee that helped bring this whole thing together it wasn't me and it wasn't Sandy it was this remarkable group of people and we'll all come up together and talk to you but Brian and Shana were challenged from the very beginning all the way through all these writing halls and all these resources all the sponsors of our sketch so thanks for helping bring this whole thing together and would you please educate us now about how to identify what this AI stuff is and how we use it with the law we have a whole video thank you alright, well thank you all for the opportunity to present here today we are the co-leaders and co-founders of the Watson legal practice at IBM and we focus not so much on practice of law use cases with regard to AI but on business of law and the sign of what we mean by that and today we're going to talk about what AI is very briefly in legal context and in the context of cognitive computing which is IBM's sliver we'll talk about blockchain and the confluence of blockchain and AI and why these two technologies have thus far been presented as separate and equal models and that's going to come to an end and how to evaluate a set of intellectual tools for evaluating what is an area do you want to say anything let me grab the clicker so we can actually start the next slide ok maybe maybe not ok we'll go along the sides so I might take it just a second and talk about how Brian and I founded Watson Legal at IBM so about five years ago I know that you're probably doing the same thing many of you in the room probably saw Watson win Jeopardy I was sitting there watching Jeopardy going ok this is super cool I was with my preteen daughter and she was really excited about Watson winning and I looked at her I said you know I'm going to go work at IBM and build robots and she kind of rolled her eyes at me because she's preteen and she was walking off well within a few months I had already interviewed with IBM and I was hired and I was excited to meet my vice president I work for now and when I first met him my first words out of my mouth after they asked me who I was was legal and he looked at me and he said what is that so as we started to form a practice Brian had the same idea Brian was one group of the company I was with and we formed our ideas together and we have now a team of many building AI solutions for a journey's road line it's a lot of fun you can't tell with the color that is a library so let's begin with what's one of the most significant promises or perhaps hopes associated with artificial intelligence looking to the near-term future we think that one of those promises one of those hopes is the ability of a machine to achieve subject matter expertise in multiple fields of knowledge simultaneous multiple legal practice areas for example and to identify unanticipated trends and correlations that produce new discoveries and new efficiencies I don't think the importance of this can be overstated suppose a machine could achieve simultaneous expertise in tax law accounting finance, economics, behavioral economics, evolutionary psychology etc what would it see what would it tell us when it was able to read broadly and understand in context information across those various domains today human intelligence is limited to the street areas of specialization and most accomplished individuals among us except for probably people here at MIT they can only be an expert in one or two fields but how would a machine consider a problem if it could consider it the lens of many different areas of expertise simultaneously perhaps thousands of different areas of expertise talk a little bit about silos here so like most organizations, law firms in corporate law departments have vulcanized IT environments adding AI to e-discovery with research, management many of those different processes is a step in the right direction but unless the data in these different systems can be analyzed simultaneously the promise of AI will go unfulfilled in other words, just including AI an existing solution is definitely not enough so we find an idea that blockchain is the fundamental promise to AI we will discuss today how and the just a simple initial framework for how to evaluate AI in blockchain use cases and such that they are having a greater chance for success we'll talk a little bit about all the workshops we've done but we have kind of figured out a methodology to a bit of what could be a bad mess but first let's qualify our terms and bear with us this is as technical as we'll get and as basic as we'll get but we do feel we need to get a little technical and little basic so we have a common foundation for the rest of our conversation so the core of our business review contracts transactions and the records of them are among the defining structures of our economic legal political systems they protect assets establish and verify identities and chronicle events today economic activity takes place on business networks in other words networks of companies that work together to accomplish certain common goals to think about buyers, sellers contractors and subcontractors in a commercial real estate project they come together in the market place where they exercise their rights and entitlements to assets and that's the ownership of transfers of the transactions of the trade value in a business network transactions involve various participants buyers, sellers etc bear with me whose business agreements and contracts are recorded in ledgers every business keeps its own records which would be many different ledgers across both individual businesses and of course entire business networks is incredibly messy reconciling those those records as time consuming prone to error, discrepancy resulted disputes and pity for very expensive intermediaries blockchain has already been by Christian we will repeat what he said but it's a distributed ledger technology allowing any participant in the business network to see the system of record a meta system of record if you will and it's radical because it's radical because it facilitates a single source of the truth potentially so if we take this and break it down into several key concepts Christian actually provided you some great definitions and we'd like to go ahead and continue on with those definitions and clarify some of the things we have up here on the slide smart contracts for example and I'll read off the definition just so that we have a good basis as we go into our one comparison a little bit later a smart contract is a piece of code that is stored in the blockchain network on the participant's database it defines the conditions to which all parties should agree certain actions are automatically executed without intermediaries so in today as we talk about privacy the history of transactions essentially a transaction log is shared between participants and is actually recorded in a sequential chain of hash linked blocks this is a mutable data storage and a couple of the words you'll hear a lot today is when we again as we continue to talk a little bit about hash but then I've heard a mutable I think probably about 20 times already this morning and then lastly we have consensus so consensus ensures that the shared ledgers are exact copies and lowers the risk of fraudulent transactions tampering would have to cross play back the same time but now they're in cognitive computing so every cognitive use case has four elements in common they understand they reason while the technologies that facilitate the use cases they learn and they interact when we say that these tools understand we mean that yes they understand structured data data-inflated relational databases they also understand unstructured data images voice files or audio but primarily narrative text the words on a page in a book or a contract or a judge's order they understand what we mean by what we say and traditionally this has been profoundly difficult for computers which could only take us at face value except for using bags of words they would only understand the literal meaning of the word and even as they improved upon that ability they still act context and what I mean by what I say in the context of a given practice area in the context of my industry in the context of my company's dealings with my company's customers and that elevates these technologies to a degree that just was not possible before certainly not to do it with human body with warm human bodies at scale these tools also read reason they put forth evidence-weighted recommendations for action based upon hypotheses that they form behind the scenes and there are thousands of ranking and scoring algorithms working to vet these hypotheses until a recommendation or an insight comes to the fore and these tools learn so that you can interact with that insight if you will and provide feedback into the tool allowing it to become more contextual allowing it to become an organic extension of a business going forward all of these capabilities depend upon large volumes of data except perhaps for interaction the concept here is less related to blockchain more related to cognitive and Watson the idea that until now we've approached computers and machines on their terms the difference being that now we can interact with them they understand us and they're able to more than in the past facilitate a relationship us and them where they approach us usually on our terms or at least we can speak more along the same lines as we've been able to in the past so this actually may be the answer to what Christian was talking about in searching blockchain AI may be the answer so many of you have probably heard that more information was created in the past two years than during all of recorded history but did you know that in 10 years the amount of the world's data will double every 12 hours so as we see here on the slide we see right now in the past just one minute 4,100,000 videos have been viewed I think most of those are by preteens 3,300,000 posts on Facebook I don't know how many of you have actually posted since we've been here I think I see some tweets going on I've tweeted a couple times so if you haven't yet feel free to go ahead and do hashtag as you're here but every 60 seconds it is just tremendously gaining momentum and as we see in 10 years every 12 hours it's going to double with more companies and law firms turning to blockchain and more data stored in distributed ledgers there's a need for advanced analysis methods which is where AI comes into the picture blockchain can help us verify, execute and record AI can help us understand, reason and learn to identify metatrends in the same way that an abundance of pixels creates a higher resolution image and reveals secondary images we may not have noticed at the lower resolution AI loves data the more data the better the model also AI only as good as the data that shows it AI leverages blockchain technology represents AI with provenance AI that actually can't trust AI and blockchain represent an entirely new paradigm comprehending the metacontext all the secondary images being able to understand the nuance of language being able to understand the nuance of whatever it is that you're trying to analyze the connections between them recognizing signals versus noise and evaluating more possibilities than you could otherwise evaluate you can see how this idea, this notion of vast amounts of information intelligently analyzed to synthesize new insights which were discussed at the beginning can be it's no panacea, but at least meaningfully facilitated blockchain technology so next we're going to cover what standards can an organization use to evaluate these technologies so as I've mentioned Watson legal was formed within the past few years by Brian and I and we focus on augmenting the business of law rather than the practice of law this was a strategic choice and honestly not one that we made this was made by our clients once we give our clients a framework for evaluating the most impactful application of cognitive computing as judged by business benefit they were the ones that prioritized the business of law use cases over the practice of law we've conducted over 100 different use case development workshops with corporate legal departments law firms and government entities in multiple countries I know this year you've been in China I've been in Russia you've been in the UK many of the major countries in the last year at each workshop we identified 10 use cases then prioritized them based on our clients goals and our technological capabilities intersected to produce the maximum business benefits we ended each workshop with one single prioritized use case so we then went back to the lab and ranked all the use cases from all workshops globally the top ranked AI and blockchain use cases had the following common sense characteristics blockchain perspective at a high level you should consider investigating a blockchain use case in greater detail and this is just initial this is just foundational if a business network is involved it can be an internal or an external business network and the answer to at least one of the following questions is yes is consensus used or should consensus be used to validate transactions is an audit trail or provenance required should it be must the record of transactions be immutable or tamper proof and should dispute resolution be final these are some of the first level considerations and there are many more and more nuanced considerations once a use case passes this first baseline test what about artificial intelligence what about cognitive computing the best use cases address a clearly recognized business opportunity or pain point back to the hammer looking for a nail scenario there are so many technologies out there that are or claimed to be AI which itself of course has no standard agreed upon definition the best chance of success that you have if you're willing to explore this technology in the relatively early days of it particularly in a legal context is to align it with something that matters to you something that's already top of mind to you and it's associated with a complex high-tivot process an example these use cases have a cognitive element understand and learn they do things that people have tended to do analyzing and surfacing insights but then weighing them against others not necessarily taking action though we don't want to be accused of the unauthorized practice of law just another reason we focus on business of law applications and they use they use the data that's available you know and the content sources of that data are likewise available and relatively easy to get to or to license if the data is third party one of our top-ranked use cases is a reduction of outside council spend we created a solution called outside council insights and we're using Watson to we went through this process Watson technology to read the narrative the unstructured portion the one item level where ten different lawyers can describe a similar billable event in ten different ways but volume is important so one of the highest volume areas of litigation or within a certain company and then how accessible is that data invoices are very accessible for me billing tools and do they have a cognitive element well yeah I mean for large P and C carriers you know analyzing or in-house departments analyzing 8 to 14,000 invoices a month they don't have a complete picture of a life cycle of a case or across cases so what are the meta trends and that's an example of a use case running through that analysis so and then we actually took it a level deeper and use the following metrics to rank each of the use cases the first one being client organizational benefit we look at them medium and low impacts to business the next one was end user benefit then we had strategic alignment and then also speed to implementation speed to implementation it could go back so some organizations have an appetite to do something transformative other organizations don't they prefer a quick win and that their vendor earn the right to move forward from there we've seen the spectrum but one of the things that we've paid attention to or tried to pay attention to is patterns of interest and feasibility from technical perspective tend to cluster how do you identify what can be done with cognitive in the legal space some of it no one's ever done before some applications are very straightforward and obvious others perhaps less so speed implementation end user benefit all of this is really about wrapping the technology around a business's needs alluring to them rather than approaching them with a technology to solve a technology problem they have a business problem we tailor these technologies to our customers business needs the language that's spoken the ontology that the tool learns that context that we mentioned earlier in one insurance company it's going to be different than in another insurance company never before and this is not just IBM but never before have has there been a scenario where vendors are able to deliver such value to clients but only if you engage with us and do more front end due diligence us and those like us and that's going to shift the relationship and the dynamic a little bit rather than showing up with a product in a box it truly is a partnership so an example of a use case that we've been considering from a blockchain perspective and a cognitive perspective is conflict checking and conflict clearing near universal need in north america when it comes to law firms not just law firms and right now they rely on reference data to determine if there are any relationships involving the firms current or past legal matters that are adverse to a client in a new matter the data that helps them make that determination is usually stored in a digital system just names no intelligence no ability to understand the various or learn the various ways that the name of an entity can be described based upon context if it's not clear a subsidiary lawyer not only that but but these tools pardon me I need glasses I'm getting old global view of every contacts contact the network how does it break down? each participant would maintain their own data set within a blockchain network now this could be internal this could be for example a large law firm with offices throughout the united states and perhaps other countries associates move very frequently now more thematic than ever before or it could be grandiose we would create a single view of the entire data set verified attorney and client identity and any modifications that would come later any updates would be stored available in a permission basis interested parties and we'd use cognitive on the blockchain we'd use cognitive on blockchain to learn and analyze variations on entity names attorneys, clients, companies relations we're doing this right now with anti-money laundering where the legacy technology right now is something like 80% false positives and a lot of that turns around identifying variations on a theme of entities ultimately identity is what we're talking about the benefits would be some of the obvious a consolidated consistent data set that reduces errors but near real-time reference data yes but improved client acquisition speed reduced disqualification reduced malpractice reduced reputational harm so blockchain will help unlock the fundamental benefits of AI and AI in turn will help make sense of a galaxy of of data that's inherently associated with blockchain technology and it's important to explore these technologies broadly and creatively but not so but not so creatively it's also important to have some kind of structure going into this process so that we're not flailing it's also important to understand that these technologies many of them can deliver value right now actually and they're already doing so if you're law firm attorneys in the industries that your clients that your clients come from in some cases in very significant ways so we need to explore and we need to explore with structure getting a sense of how we frame the practical utility of these technologies at a very fundamental level will help us achieve a place where we're able to derive value from these technologies sooner rather than later so you've begun that exploration here we all have together today and we look forward to spending the next two days with you as you evaluate this technology and as you begin to think about it in the context of your own firms and as you bring and organizations as you bring us ideas and challenge us let's challenge each other so thank you for your time alright see I promised you the good stuff alright so questions, comments ideas alright James introduce yourself James Miller I'm curious if you could answer sort of an applied question sort of an applied problem that you've seen sort of in the practice of architecting these solutions one of the things that I think troubles a lot of practitioners and people sitting at a desk trying to get worked on is platforms are largely legacy in a lot of environments they're not in a form that you know and you sort of touched on the idea that you know sort of the data back end and trying to piece things together certainly there's ways that people do that effectively MuleSoft and other things I think are great products but do sort of take the perspective that you have to start with a data warehouse and sort of a modern sort of data centric IT sort of structure or can you can you move from an application sort of oriented environment how do you cobble these together and what sort of technologies are more amenable to sort of a more hybrid solution than something that's like you know throw out and then rebuild everything with new home deco can I out you a little bit look your context is attorney is it at FCC FCC is great work he's also one of the great legal hackers in legal hackers group and it's data science at the legal hackers global summit in Brooklyn last summer so he's coming from that perspective but the thing I would go in from that is well wow big difficult enterprise complex federal enterprise of a regulatory agency that's I think partly why he's asking well I think the complexities is one of the reasons we started out with kind of a simple boring use case we're working with invoicing and bills that is not what Brian and I came in thinking that we're actually going to be doing when we had when we went and had well we started out with a few workshops and we ended up with over a hundred because we were really trying to figure out how to move the needle not only just in the way that you had just mentioned being able to manage so much data that we can go after we really had to focus down in on what was accessible at this moment and voicing was easy and when we did we were connecting to discrete data sources yeah you know yes I think that if you're dealing with a kind of meta context or meta trend use case absolutely you require a data lake data mark approach but otherwise and the majority of the time we're still connecting to discrete sources yeah and it's and that may instead of boiling the ocean I think that's one of the things that we would recommend is start to look at some of these smaller use cases they may not be as sexy they may not be as fun but it's a good way to kind of jump in so you're not jumping in the deep end well and to qualify it's not sexier fun but well I disagree I actually think it's pretty fun you think voicing is fun well here's what I think is fun so you set this up so perfectly so we didn't so the industry standard savings for competitive tools 4.5% reduction in outside counsel spend we're doing 20 to 33% so that is where I guess if you think a $392 million savings is sexy in that case yeah I guess it is I never thought about it that way proving the sound sober business value nobody's been so excited I didn't see any of that no so yeah I think it's moving the needle in little ways instead of jumping in and building your first robot yeah yeah next question thank you I don't think I was actually first with the hand up but I was closest that was like that my name's Pip Ryan I'm a barrister from Sydney and I'm also a lecturer at the University of Technology and I loved your presentation thank you you sounded as though you were a little sorry that it was the business of law that had been chosen for the work as opposed to practice but I was listening to your presentation thinking I would be terrified to try and do what you're doing on the practice of law do you feel the same way now or are you champing at the bit to go to that as well well maybe we have different opinions how do you feel I think we actually will probably agree a bit on this one you know I would love to jump into the practice of law because I think that our citizens in so many cases could really use artificial artificial intelligence and also the transparency that blockchain brings us 80% of our citizens in the United States do not have the ability to have legal counsel because they can afford it and so I would love an opportunity to help them with the practice of law and with some of those simple things like building forms you know some of the things that we've seen by you know some of the AI that's been developed Joshua a student from Stanford has created do not pay assisting with UK citizens with over 5 million in savings on their what is it traffic bill so there's a lot of good use cases and options for that I think on our end we needed to kind of move that needle on return of investment and invoicing was a great way to do it volume again so one of the practice of law use case we've considered them and we are doing some of them actually but we often ask what is what are the highest volume areas of litigation for example in a given geo or in a given practice area that way there's the broadest possible value that we can deliver to the broadest number of clients who are relevant it also drives down costs but it has the benefit of there being just a tremendous amount of data to see nuances and variations on the theme of scenarios there is more of that kind of information in business of law applications or business of scenarios it so happens than in practice of law scenarios well in a great place that we found in the invoices within the narrative there's so much good juicy information for Watson to learn that as Watson starts to really understand the nuances of the languages within the invoices and specifically within the narrative it's a great place for Watson to start with that learning and then we can add at that point pleadings we can add motions we can add the rest of the bodies of law for Watson to then learn so thank you Hi my name is Anya Titova I'm a legal design resident at IDEO and I was a litigator before that so I kind of have a very specific question about your conflicts check scenario that you presented and the litigator in me kind of gets nervous and wants to know sort of how do you de-risk exposing existing conflicts in a law from database sort of the law on conflicts are so fuzzy and still developing especially with respect to what is consensual waiver and things like that so you know things are known to sort of be very risk averse so how do you know when is the right time and how do you convince lawyers that it's okay to deploy this technology without sort of opening this Pandora's box of existing issues I think the answer to that is so this is a use case that we put forth to bring here to spark a conversation one of the lightning conversations that we might have tomorrow we're looking for your feedback so this is a volunteer use case that we want to explore together to the degree that that's fuzzy and gray we don't have an answer to that question we'd like the advice of lawyers actually this is an example ideally really of vendors and practitioners and scholars working together yeah so we're trying to tame the wild west we'll say yes sir or whoever has the oh yes I'm Jules I'm from Lunicat Ventures we're trying in the most exciting legal tech companies out there and I know that Ross and some of the other entrepreneurs out there are licensing Watson are you seeing more of that and if not or even if so what are some of the areas that should be using Watson for entrepreneurs to create kind of more radical solutions we have a list as we have over 100 information from 100 workshops so we certainly have with Watson which is an approach where we license the technology the building blocks to those clients or customers they build solutions for themselves and then our business our practice focuses on actually going out and building and designing those solutions and selling them into the marketplace and we're able to well of course we're able to take advantage of IBM's scale to help with R&D frankly so we're seeing both we are seeing a lot around insurance and law related insurance use cases and I don't know how specific I can get without giving too much away but I'll tell you one and it's pretty common sense I don't think it's giving anything away which is the cost to settle a lawsuit is to find another line of claims is three times more expensive than the cost to settle the same dispute when it was a claim so early predicting which claims are likely to become lawsuits by using cognitive to understand to read settlement agreements to read initial pleadings to read claims files etc is a potential use case here and there are many different ways you could go about scoping that use case well an AI differs depending on what is built it excuse me we spend about five billion dollars every year in R&D and so the Watson that we get to play with is pretty cool the Watson that some of the companies have maybe put into place earlier on are some of the versions of Watson Watson right now is over what 60 different things everything from building blocks we can kind of go through those a little bit later if anyone has any further questions but yeah there's definitely a lot of good companies that are using Watson using other AI tools to really answer the needs that we see in the pain points in the marketplace so and that they see yeah yeah I think we've got one more one okay very good my name is Yontan Lawrence my background is in language modeling and and I'm a little bit curious about what you mentioned in your description of cognitive systems with respect to IBM you sort of had these four categories understanding, reasoning, learning and interaction and I'm a little curious about understanding so I think one of the biggest concerns with AI is trust and accountability and I'm curious how you guys measure understanding by the business result we go in trying to we go in trying to help a client achieve a certain very specific business goal and if the result achieves that and there's a baseline and it achieves it by X percentage over that baseline that we define with the client on the front end in terms of what the criteria for success are it's very idiosyncratic in other words so a lot of it comes down to the first standing Watson has to be trained Watson when you get it out of the package is about five and so as we like to say Watson's currently going to law school Watson just learned all the financial services regulations in the medical area Watson has graduated medical school and has read what 23 million journals or something exotic like that so we're doing that also in the legal space as Watson's starting to head down that path of hopefully getting his law degree soon so but yeah that has a lot to do with the understanding because he can't understand what he hasn't been taught yeah okay any it's now or later are we good sure okay so thank you thank you Brian IBM Watson can I grab one of your libeliers whoever is easiest to get okay thanks so now I'm going to ask so we've got one more thing to do before a sumptuous lunch and that is sorry I disappeared down here because I have to monkey with the slides a little bit let me pull this up so this is oops boom so we're calling it lightning talks on the breakouts and the workshops so what's going to happen next I'm going to ask Brian if you'd come up and and everybody that's doing a breakout session Tony, Pip all y'all come on up, definitely Michael and for the most part what you're going to hear is like like a one or two minute lightning blurb of like this is what we're talking about all of them should start with a fundamental area of law or I might even interrupt but it just needs to be like bankruptcy talking about blockchain and asset registry property encapsulating the value of a carbon credit supply chain et cetera so that's what we're listening for and then get a feel for the topics who's talking what discussion you'd be interested in going to talk to people over lunch if you're interested to follow up we're all going to be going to lunch right after the lightning talks and that's the intention is to socialize that a little bit think about it and then you'll select yourselves into the breakouts in the far room we've got 11 banquet tables that sustain about nine people each and we found that's about the maximum that can sustain a meaningful dialogue and four of those tables have monitors and so if any of you really feel that you need slides you need to be looking at stuff as part of what you're doing I'd say go and grab the monitor table first or talk to me or one of the organizers and as a reminder we're going to ask that somebody in addition to these facilitators ideally that isn't the facilitator also take note so they might try to nominate somebody in the session you could also do it as a shared thing and we're putting all those notes into the same slide deck as I mentioned earlier alrighty so to get started why don't we start with Tony and Pip and why don't you introduce yourselves and just say real quick outline what your breakout session is about Hi everyone my name is Tony Lai I'm the CEO and co-founder of a company called Legal.io we're building infrastructure to make legal services more accessible Hi I'm Dr. Pip Bryan and my PhD explored the liability of third parties for breach of trust I've also been practicing at the bar for 15 years and I'm on the International Standards Organization Smart Contracts Working Group and on the Standards Australia Blockchain Technical Committee again heading up smart contracts so as a lawyer and a legal technologist one of the things I'm confronted by is how I can ethically introduce AI into my practice this is a a cause that I share with many of the lawyers all across the country who I've been talking to who are keen to try and utilize technology in an ethical way as part of delivering their legal services now as Shauna and Bryan pointed out there is the specter of unlawful practice of law and regulators all across the country all across the world are in fact looking at the ways in which lawyers can ethically in compliance with their fiduciary duties help clients in a way that gives due understanding and a level of responsibility to their duties as lawyers to look out for the true interests of the client and to what extent this may come into conflict with an AI system that they don't have full accessibility into the workings of and so Pip's going to describe here one of the scenarios that we're going to go through in terms of trying to address some of these legal ethics issues when it comes to how you can apply AI Okay, so speaking to Shauna's suggestion that the reason why she would like to go into the practice of law with Watson Legal which is to tap the unbanked or the unlawyered market so this is the group who aren't so poor that they can get legal aid and they're not so rich that they can afford a lawyer so people like me and so what you want is for this middle group who want a lawyer you want to be able to provide services to them so you set up a law firm and what you decide to do is automate some of the advice there are two arms to this the first one is the fully automated robo-advice so your robot is using artificial intelligence to inform how it's going to advise and it understands the advice that it's being asked to give because of voice recognition technology all powered by natural processing language services and informed by artificial intelligence the problem this particular group has is they've got to get underneath the hood of the data and also the people programming how that data is then going to be filtered so if it was Watson Legal for example, you'd want to get under the hood and know exactly how that was being driven so that you can defend it in court because your primary duty is to the court and then to your client so we've set up the scenario we're going to pose those questions we want to be lawyers in that room but we want to talk about blockchain so the second is this correct? I don't think we had it on the sheet but it's legal ethics scenario so that's the legal ethics scenario and hit the second one what's it called so I can pull it up? so the second one we're going to call ownership of personal data that's the area of law and what we want to look at is not necessarily just self-sovereignty but the idea that there are organizations that are collecting information they're collecting personal data some of that may be private personal information it could be metadata about your movements as you use your handheld devices that are internet enabled it could be your medical records your correspondence on social media and the important thing to note here is what we want to explore in this particular scenario is sometimes you're creating the data about yourself your post on Facebook, your email sometimes other people are creating it about you your medical records or they're tracking your metadata the question is who owns it what are the obligations and what can you do with it what can you not do with it of course this is going to be jurisdiction specific but we'll build that into the discussion great and that's going to be day 2 we think and good fit for the identity stuff alright so thank you Pip and Tony so if you're interested in that mingle to chat with them over lunch and go to their session right behind me here alright so next up Bob Craig who's also an honored member of our program committee hearing more from us before the lunch talk and who's agreed to do some heavy lifting on the scenarios bankruptcy it's particularly heavy lifting because I'm not a lawyer the CIO for Baker and Hosteller have been for more than 20 years now in the legal industry for 26 out of a career that's longer than I care to mention in any event the bankruptcy scenario is really a bit of a thought experiment so I'd encourage those lawyers who are here if you want to explore some of the practical possibilities of where blockchain could come to bear on a complex process so the idea essentially was to find a scenario that had lots of parties involved these parties don't particularly trust each other in the context of a bankruptcy but they have to trust the outcome of the process so you have lots of legal entities involved making claims, you have lots of lawyers representing those legal entities you have a bankruptcy court and then of course you have the bankruptcy trustee and so we'll explore from a couple different dimensions where blockchain might be able to come to bear on that process I just thought it'd be awesome put it on a blockchain but Christian kind of blew that all up so we're going to actually have to think so the good news is I've loyered up I have two terrific lawyers who are going to kind of co-moderate Oren Warshowski is a partner with Bakron Hostetler Nina and we have Nina Kilbride Nina and Oren show yourselves come on because otherwise people won't know what you look like and who to talk to sorry Oren, I was going to try now Oren is a partner at Bakron Hostetler he has been intimately involved in our representation of SIPA and the made-off matter so he's got some interesting bankruptcy scenarios to bring to bear and Oren will focus on kind of the process of the bankruptcy proceeding as a scenario Nina Kilbride is a director of legal engineering at Monax, has some really innovative thoughts around the whole notion of legal engineering as it relates to smart contracts and blockchain and Nina's scenario is really dealing with a fact pattern that involves cryptocurrency in the context of a bankruptcy which I think is fascinating so I think I'll leave it at that it should be fun and illuminating, thanks Bob bankruptcy you'd think with the blockchain this could all go a lot better so let's do the math and see how Brian, Thompson Reuters Labs yeah hi, so my name is Brian Ulyssany I run the Thompson Reuters Labs in Boston we're one of seven labs globally so I'm neither a lawyer nor a tax accountant coming from our tax and accounting business so we of course provide financial legal tax and news information to professionals worldwide but this is an interesting scenario that I wanted to talk about from Richard Ainsworth at BU Law I don't know if Richard's here today no but he's basically set up an interesting scenario for a proposed architecture for doing setting up an architecture for VAT tax collection so I'm using blockchain and AI and proposes this for the Gulf Coast which is going to a VAT tax and so the idea here is basically that as you know in VAT tax is as goods are enhanced, materials are enhanced and the value of a product is increased you pay taxes on the difference between what you received and what you produce and currently the VAT tax regime in Europe and so on allows for, enables a lot of fraud so you can frauds that you can commit by sort of delaying your reporting or not reporting correctly and then you can collect monies that you don't you're not entitled to Ainsworth's scenario here is illustrated by a number of manufacturers here these blocks at the bottom ABCD there are various manufacturers and businesses across the Gulf Coast and they're producing these goods and services and these transactions are being monitored and sort of adjudicated by these nodes that form the consensus protocol for the blockchain and they're voting on using AI and they're saying we think each node will say you know I think this transaction is legit and should be and the VAT that should be applicable is such and such amount when the consensus is reached and enough of these nodes agree on the VAT analysis then it's put on to the blockchain and Ainsworth's suggestion is that the number of nodes that are distributed here should be proportional to the GDP of the countries represented so the purpose of the scenario of the session that I'm going to be leading is to kick the tires on this architectures see what people thought see what sorts of risks and opportunities you see here and think about whether this is the way to go great thank you Brian VAT very mechanistic seems like a good fit for something computational Chris if you could introduce yourself and your scenario and your breakout my name is Chris Jaggers I'm the CEO of Learning Machine our topic will be official records that are owned by people and directly used in the world our company works with school systems governments and companies to issue official records to people in a format in a digital format that they can own and use directly when applying for subsequent education or jobs we co-chair the W3C's credential community group and you know this brings up interesting topics of you know cross cross jurisdictional issues when you're using an official record that's relying on decentralized verification of its authenticity you have to use the word diploma and MIT so even though the technology can support any kind of official record one example that just came out is MIT began issuing their diplomas doing this on the bitcoin on the bitcoin blockchain that is inscribed with the students public key so they can prove ownership of it signed by MIT's private key so the provenance can be demonstrated and the students from the initial groups of the MIT Media Lab and Sloan School of Business they can use these documents anywhere in the world and rely on decentralized verification and MIT is going to be expanding it to more groups as we go on that's just one example of many yeah thank you so if you want to talk about that more join me thank you thank you yeah I want to go to thank you so this is a good example can you think of official records that come up in your life and your business this is interesting okay who wants to go next Harrison do you want to start with the work do you want to start with the work so why don't you start with your small session then do the workshop and yeah I guess I'll say okay I love the idea of like the star starting team all up at once but you can go and sit down after your talk it's alright if you introduce yourself in your team and the um link the automated legal entity one got it hi everyone my name is Harrison Pearl and I am founder of C4 coin here with my team and we are going to be leading an online session today so it'll be at 115 and we're going to be exploring what does it mean to automate a legal entity so the area of law is uh partnership law uh corporate law and understanding what would it mean to give officer roles to a computer uh what rights would that computer have can they incorporate their own entities and then sue and be sued so that that will be the small online session that we're leading uh my background is in accounting so I'm not a lawyer uh come from an entrepreneurial and accounting perspective on on these topics so we're going to get a little bit gappy a little bit IFRSE kind of get real up in here um thank you very much gappy um and so actually not so quick um so come on up um is dan harpal here dan harpal okay yeah so and now um there's a second um thing happening in we're it's going to be a workshop a little different format um a little bigger um won't be limited to people around the table we'll do it in the big classroom across the way and um michael casey is going to be leading that with with a couple of other superstars so michael yeah thanks for I think you most of you met me earlier um uh haristan setting up quite nicely because the project that we're working on to some extent is going to require understandings of the roles that machines and the rights that machines play in this in this circumstance I am definitely not a lawyer uh and therefore I'm looking upon this I'll be completely frank as an opportunity to get a little bit of pro bono legal advice um we have we think a very exciting uh research project uh underway and and and hope for you will agree that it's worthy of pro bono advice and that is uh we're looking at some of the devastated regions of the caribbean right now where there is as you would know literally no grid and thinking about whether this is the right opportunity to go into them and experiment with an idea that many have been talking about for some time this notion of energy democracy that we could actually rebuild the grid from scratch to really emphasize what many people now see as a great value proposition uh a decentralized architecture not only because of the resiliency that a decentralized grid would would give rather than these vulnerable centralized systems but also because um empowering users within the system uh with smart devices and the capacity to decide when and what to develop in terms of generation capacity their own solar generators and so forth and then trading it amongst themselves in an open price driven market environment should extract we would think based on standard market economics uh so much more efficiency from our energy usage and energy construction in a very constructive way so we're going to experiment and go down to these places and set these things up but it obviously raises a whole host of interesting legal questions um how do we represent rights to ownership of that grid because this grid may or may not come from just ganging together homes solar panels it might be that there's collective ownership in a common community grid um how would we uh could be collateralized those claims in some way how would we represent them um and then you know with Harrison's help through C4 coins interest in carbon credits and these sorts of things are their ways in which we can take this into some sort of international realm there's questions about the relationship between the public utilities these legacy institutions around which regulation has been traditionally developed uh how do we interface with them should we um so ultimately we're looking to find a way to really design a different looking electricity grid based on this technology a smart grid uh and um and how do we uh how do we you know make it legal did you want to add something to that? so so again this this will be about uh Michael's project of a distributed solar micro grid and then Dan Harpel and I will be coming in Dan Harpel is going to talk a little bit about uh maintaining data provenance or data veracity from an off-chain source and bringing it onto a blockchain and I'll talk a little bit about the difference between private and public blockchain implementations and what issues those bring up in terms of token and network ownership the interface between those private and public domains in this particular structure right yeah I think there may be I mean public utilities commission you know consumer protection like it's this is like um skeet shooting target practice for issue spotting so um please come spot issues it's going to be a workshop format that um we are good at if I could say and Michael's good at in the media lab it'll go a little bit longer and we'll really take we'll workshop it and see and then what will happen next is you're going to go to the Caribbean and do it so join us and learn um oh great come on um who else so Cassius law who else is doing I just mentioned a few people to get people up here to get the idea that where it's time to talk who else is supposed to be up here that is an up here now that hasn't talked it but you think you're talking you want to be talking anybody okay so now is the time all right so um let's do grvinda next um and then we'll go and then we'll go right down the line here you go I'll pop your slides grvinda all right thank you um so as the slides are coming up uh just a few warm up comments I you know I was on the stage uh just a few months ago previously previously in my IBM persona and uh one of the slides uh can I just uh chop it forward uh slide show whatever it is is one more is uh you know I like to invoke the work of ronal course very often it provides a very good description to put uh all the all the all the mess that we see uh around technology into into the proper a proper perspective and a few months ago I had only maybe the first one two three icons and uh preparation to step the next little icons over there the idea being how the nature of phones and other entities is being reordered as a result of now uh chain which is just the most recent technology kind of disruption that we are seeing and what is going to be the nature of the firm in this 100 years from 1937 to now or to to 2037 if you project forward what are the possible entities of the firm and how does that test out the theory of ronal course uh around phones and markets and law as well as you know existing frameworks of law so if you move to the next one so I go into supply chain I've been doing supply chain work for quite some time specifically or more genetically blockchain for the last about four five years the work was uh with the core team at IBM previously and that was uh standing up the blockchain use case incidentally on IoT so for those of you who are familiar with the IoT space it ties actually very closely with the supply chain space one way I look at it and I'd like you also to look at it is from the perspective of three floors the physical floor there's a digital floor and there's a digital floor part of the magic and part of the challenge for all of us is how do we converse these three floors it's quite a mess right now so how do we bring those three floors together and in the in the mix of that there's also all these new entity forms that are so problems before you stop previewing yeah so so provenance is one of the burning problems uh around supply chain right other problems are dispute resolution uh double spend that was being discussed around and how to avoid that provenance specifically is one of those big loaded terms what do we mean about what do we mean with what do we mean by it is what I want to unpack in the session later today and specifically what I want to do is kind of debunk the myth that seems to be floating around that if you put a blockchain network you automatically will have provenance right so we want to get rid of that myth but then that begs the question what else what are the elements of provenance if you're able to move to the next slide Daza so what are the elements of provenance and I want to look at this framework a little bit more depth I'm not going to say this is correct I'm not going to say this is exhaustive but I want to get the juices flowing sufficiently that we can improve on certain structures and certain frameworks like this provenance at a minimum means origin it means access it means custody all the terms that the lawyers can better speak about it means jurisdiction it means possession so on and so forth some of those icons I've tried to represent and provenance doesn't mean just tracking it's one big element of it but tracking and tagging technologies are really not new so what else does it mean is kind of what we want to cover it should be pretty clear by now I don't come from the lawyer side of the house I come from the systems and architecture passion of the house that's where I can't look from but yes I do come from a family of lawyers and one of them my wife is right here and we've got a few power couples here where it's lawyer engineer you know like the philosopher kings of the next generation so your vendors is one of them and it's great when you both come there's multiple of these so thank you your vendor while you're kind of blowing me away the slides have come a long way so there you go supply chain go vendor so we've got a couple more let's do James so again thanks for the introduction earlier he mentioned some affiliation with the FCC I am an attorney there I'm not speaking on behalf of the federal government so don't get confused and think that all my mistakes are attributable to anything in the administration they have their own we're going to be talking today in the public sector session at 115 mostly about rulemaking and I think what we'll observe are that in many ways the same sorts of stresses that we see in other sectors with scale and the scope of problems expanding big data those sorts of things we see those stressing how government functions and so in the public sector we have a number of particular constraints and particular sorts of trust problems and things like that we'll be talking through maybe how some of these technologies can are in some ways both for creating new problems but then also are obviously things that we can leverage to help fix them so I hope you can mix in some of the trust and transparency and maybe there are privacy issues and all sorts of other stuff but it should be a fun section talking about public sector stuff rulemaking super exciting even more exciting than bankruptcy it is do you want to talk about tomorrow also for pedagogy and data science? a quick squib he mentioned that we had been looking at what is the legal hacker both identifying what they do what sort of pedagogy sort of strategies can help to create people who are modern information workers in the space where law and computer science have converged we'll talk a little bit about that maybe part more in the workshop maybe do some ETL kind of things I used to be more of a localization network guy so you might see some pearl and regular expressions in that session yeah I'm going it's really great like this is where we learn so thank you James great rulemaking and it's a number of people have said by the way including sanny's been pressuring me for years now figure out how law can be made in ways that are data-driven in ways that are that can under can take sensor input by design to see how well are we performing is the law achieving its purpose is there need to adapt can we build that into the rulemaking process that's a hard question but slow progress is being made some people looking at that in a municipal context and James is particularly well suited to be able to help us make progress and rulemaking it's also called just governance of our social compact so yeah it's exciting okay Phil on litigation alright thank you does hi I'm Phil Rosenthal the president and co-founder of fast case and we also did the ASN box I actually am physicist though from Caltech I don't know if that gets me in trouble here and lawyer too but so we're going to be doing a big messy class action tort regulatory compliance problem but I want to talk about what we're trying to illustrate and and really what it is we want to try to shift the center of the universe so it's just you know small things so what I mean by that is you know that if you're in a law firm or a corporation if you're on the legal side typically all AI really all technology is read only right you get whatever software you get you can click the buttons you're told and and that's another wonderful amazing tools but we know that the litigation will show it some of the most basic questions how was what should I settle for how long it's going to take you get anecdata these days and it's it's very bizarre and the problem is that the center of the legal universe is typically the big publisher or technology company and that's where that's where the center is now it's very darkest the greatest database in the world is what you have at the firm or at the corporation because you got all that good confidential stuff the settlements the actual billing information and and yet it's not being extracted now the problem is is it's so often unstructured and that's where folks like IBM Watson come in of course and Brian Cune is going to be very honored to be joining us for this program to talk about how that how that can help with these scenarios and and so it's it's also the other problems the silos they talked about right to really get the insights you need to pull the data together public data third party data your private data so what if we can invert the model what if we can make the firm or the corporation the center of the universe and have everyone put their data deconstruct the tools put whatever is needed under the firewall right so everything can be together because that confidential data is not going anywhere else so everything else should go there what if we could do that and then it's ai is no longer read only then the folks at the firms can be the makers of ai and and make wonderful tools and this is really a community effort of all of us there's a lot of companies involved doing this because we got to get all the pieces in there so we're going to look at a night at a actually a company that was putting nuclear material and watches actually happens and and cancer claims and regulatory problems but what does that mean in terms of the beauty contest in terms of running the litigation all the questions many of many of the questions that come up and ways that we can use ai to start going about answering them with this new model and look forward to thank you great not so fast so I just had an epiphany I'm sorry if this is already known but it sounds to me like you're talking about going from artificial intelligence by I'm part of this paradigm which we love sandy and around we love of having the core data situated in a environment where it's clear what the ownership and control is we get a lot of performance and insurance capabilities and cloud services and they're great not great for everything for very sensitive high value mission critical other data maybe there's other things that are appropriate so your concept of bringing it into like a data lake or to something where it's going to control where you can go much deeper and look at the facets and the features and the dimensions of the data very interesting and maybe think okay so now what we're talking to intelligence now is attorney intelligence that's the new ai that you're starting to explore I think or it's the old wisdom applied in a new way so let's do that I'd like that pull down menu please soon okay great thank you okay thank you Phil's litigation talk to him about that and now last but not least people that in a really cool law firm that yearns to be digital and integrated not a big firm and we work we have our offices together in Cambridge Innovation Center and we sometimes talk in the kitchen about God why what's up with this mess of like all these emails many of you may have noticed especially in internet land it's hard for me to respond to all the emails it's hard for everybody in the attachments here to come through the form and is in the file why can't this be integrated why can't it be automated so before you talk about your session I want to thank you for writing such a great paper for this conference about automation and integration of law offices that I decided and we talked but we're going to make up proceedings of this conference and there's a few other people that had some conferences we'll put a call out for people through other people at the conference that have given some papers and some good like medium posts we'll put a call out and we have a Google form where you can upload them we'll allow some time afterwards but yours is in fact you sparked it so we'll send around by email at the end of the day on their paper so everyone can read it and we can start to talk about this and the vision that you have for getting out there and transitioning the law from an office perspective especially a small office in immigration practice but really all of us to that digital world that Sandy was talking about but first we have a particular topic and so your scenario is immigration right immigration so hit it so I'd like to thank Daza and really the whole community for bringing us all together I'm an immigration attorney turned Cassius I assumed that we were up here so I guess not so Rachel Cassius we I am the owner and founder of Cassius law we assist researchers entrepreneurs companies and obtaining visas permanent residence citizenship really the whole myriad of business immigration work and really what I found in my practice was the fact that I was just recreating the information with every case we have systems in place I have a form software that I use we have just tools that are not integrated and I had the wonderful really luck to find Dr. Beberman who is just a brilliant mind electrical engineer by training I did a postdoc at MIT who actually just had a successful exit from his company so I get to rob him of his brilliance but really we're looking to understand the digital transformation in law and specifically looking at the immigration perspective nice he has a big head so this is a breakthrough so based on the large number of stakeholders within the immigration process you have government entities that have a key mission to decrease fraud maintain the immigration rules and then you have companies who are just trying to navigate this maelstrom of regulatory issues but then also timing we deal with a lot of pharmaceutical companies who need minds from different countries who right now are dealing with certain executive orders that are vetting previously non-criminal people who like why are we putting so much scrutiny on researchers when clearly they're here to do good work so basically our whole project is to understand how we can bring technology and lawyering into an automated system where we can upload documents into the cloud and really what if we could just send a case to USCIS electronically and have them access it in the portal instead of printing things out that were once in pdf to then have it just get sent out by mael and then having the process then get scanned the documents getting scanned into a scanner and this is what immigration is doing every day so very outdated and importantly we're here to compel you guys to put your genetic sequence in the election so you can identify it for deportation this is the best idea but yeah thanks again thanks for everybody for organizing this I think this is a collaboration of all these different skill sets that are coming together and some of these themes are just really really great and unique probably to this conference I think as Daza and Rachel mentioned there's also beyond this there's some obvious things out there people are asking why isn't the system electronic right I mean if I can do my bank transactions electronically why can't I submit something to like a government entity right as a lawyer electronically things like that and another big part of this is there's a lot of exciting technologies that we're looking at here there's blockchains there's AI there's all kinds of stuff automation right all these things and there's already other technologies that are actually already proven have been around for at least a decade giving people positive ROI talking about cloud computing big data analytics, virtualization we talked a little bit about e-signatures and things like that the big question for the attorneys in the room that have their own firms is how does all this technology how do you personally ingest it not just understanding how the technologies work and how it fits into like a global landscape but as a firm as your own firm how do you ingest it, how do you use it how do you become the kind of leaders in this technology themselves and this is a difficult problem because some of the things that we talked about the IBM guys mentioned the silos including data silos but also silos organizationally yeah there's a lot to talk about this we're going to do these lightning talks and we're happy to engage the community and thank you for the help thanks I'm barely MIT, you're fully MIT but we do speak frankly and lovingly but frankly to one another I encourage all of us to speak frankly and come to it quickly and see if we can make progress before you go and we have the last ringer here I want to make sure it didn't escape there's a quote in your paper that really summarizes a lot of the stuff behind what you were saying it was that CIO quote from JetBlue can you say it listen to this this is a really interesting one he sees IT and digital transformation not as like a siloed skill set to some specific people in your organization but as like a global tool set that you can use everybody, right, all the stakeholders so I want to see green cards, passports on blockchain there you go, thanks so this is a great session because we can remake the law with these folks that's what we talk about in the kitchen over there we can talk about it here, circle up make progress let's do that, thank you for coming I know it's a work day and you've got a lot of things to do so I appreciate it for everybody okay, now last but not least, Don Tebow executive director of Open Identity Exchange among other things and one of our identity fissionados, the identity tomorrow's going to be primarily about identity today's more about contracts and setting the tone and so Don has a seminal use case as well that will be convening a breakout around and it's air travel, if I am not incorrect I think Don's enthusiasm was that this use case touches everybody in the room and I think most of the topics that we've talked about yet, so the International Airline Transport Association has a vision which they call one identity but the use case is each of us walk through the airport without breaking our stride so if you think about Sandy's beginning where we live in this world of digital breadcrumbs or perhaps more accurately this torrent of digital signals that we are delivering via our smartphones and surround us in an airport from a security point of view from a retail point of view from a governance and government point of view the IATA has a clear and present problem right now they need to reconstruct the identity system that under is the infrastructure for international travel so recently the Russian Federation has insisted that all personally identifiable information and passenger name records be housed in a server in the Russian Federation the Trump administration has said that at the end of 2018 we'll need a positive identity check for exit of course the US requirement cascades throughout regulatory and governmental agencies around the world where we'll have identity checks on the way in identity checks on the way out so the multi-stakeholder environment is that of airlines airport authority regulators all trying to sort out a new piece of infrastructure that will provide the governance for this business legal and technical framework our favorite that's the BLT sandwich that is extremely complex and a particularly wicked problem so the IATA has been convening workshop after workshop trying to sort out the trust framework that will guide the new kinds of technology coming through biometrics the new kinds of regulatory requirements coming from the Trump administration the EU's GDPR requirements for privacy etc so you have in this use case more personal travel and many of the issues that Daza and others have talked about so it's a real world problem the IATA folks are looking out for help and the most difficult thing that we've had is engaging legal counsel from the airlines from the governmental authorities from the airports authorities trying to create this trust framework of course the obvious solution for so much of this data and for the trust among parties is some kind of distributed ledger so we invite your input and your thoughts into the workshop that we'll be doing thank you Don Tebow open identity exchange awesome alright quick show of hands we could do a little bit of Q&A if there's any urgent questions about what's happening or anything like what happened or we could I love the sound the buzz of dialogue when Sandy needed to speak we had to start this thing we could just go right back out to the reception area have some dialogue and in about 15 minutes lunch starts make your calls meet your people and then at lunch we've got Casey Coleman also the program committee will be speaking a little bit to do some more framing and to make sure everyone's sorted out for the session so let me just ask just quick show of hands is there anyone that has any like we're a little bit early which in my book is good but are there any questions or issues or problems ideas that must be spoken now hearing none I would like I well by my authority adjourn the morning session and we'll stand in recess until basically the next session after lunch when we'll start the breakouts and then we'll have a special we'll have a special lunch speaker with Casey Coleman and what we'll ask is when we get into lunch a little bit when you get your food and your chat we're going to ask that people actually filter back in here to hear Casey unless we come up with a better idea alright so thank you everybody morning adjourned