The second Internet of Agreements conference brought together some of the best minds from areas of law, supply chain, and fintech to discuss the intersections of blockchain and global trade. Since global trade consists of cross-border agreements on top of local commercial agreements, the regulatory structures surrounding all of this commercial activity are significant and varied in complexity. The efforts to shift the accounting processes from traditional systems onto a distributed ledger such as blockchain and program executable deals via smart contracts opens up a lot questions on the role of law in a distributed financial system. On this panel mediated by Mattereum’s Chief Legal Officer Christopher Wray, forward-thinking legal minds with an interest in blockchain-mediated commerce and the role of law discussed the significant implications of this technology moving forward. What is the relationship between law and software in the context of blockchain and programmable agreements? Will our automation become so effective and error-free that written contracts become truly obsolete? What do we mean when we describe these blockchain solutions as “trustless?”
The panel highlighted the significant role of law in new methods of digital commerce which is often overlooked in the high-speed space of blockchain and cryptocurrencies.
Vinay: Would everybody like to make a tweet indicating that they’re here? The hashtag is #IofA. Make a tweet indicating that you’re here, indicating what you’re interested in and what you’re doing, and at that point people could just flick through all of those tweets using the hashtag, they could figure out who they want to talk to later on, they can see your picture of your profile, and this will help you figure out who you want to be talking to. #IofA.
Chris: I have a very esteemed panel of legal minds here this morning. I will let them introduce themselves in a moment, and then maybe have a 2-3 minute introduction, to frame the discussion this morning. But we’re going to keep it pretty open, so if you have questions as we go through, please do ask them. I won’t necessarily allow an answer at that time; it depends on salient I think your question is. So, if you’d like to just introduce yourselves in a few sentences, just who you are, and maybe just a little about your experience in this space? Stephen.
Steve: Hi, I’m Steve Palley, I’m a lawyer. You can tell by my accent that I’m from the United States. My practice, I’ve been a trial lawyer, a litigator. I started out at an interesting time in the practice of law, where you still used books and you weren’t allowed to use electronic research until you learned how to use the book. I was also the first person in my firm to have Internet access. I had to explain what an email address was and why I needed one, and I had a Mosaic browser. I was the first person to have an Internet and a web browser in my law firm.
I’ve had a very traditional practice, I spent about a month in trial against your friends here in London, Lloyd’s, in September, dealing with 40-year-old insurance policies and telexes. But at the same time, I was sort of an amateur hacker as a kid, I taught myself to program on a TRS-80, writing adventure games in BASIC. I lost that, I came back to it, and then about five years ago I ended up developing a dispute resolution platform, using Ruby on Rails and a centralised database using PostgreSQL, and I realised the problem with my database was trust.
I was like, “Other lawyers out there must be interested in software,” so I did a little googling and I ran across a fellow named Casey Coleman, whom I’m sure some of you know, and he introduced me to blockchain and I discovered Bitcoin, I think it was Ripple, actually. Independently, I was trying to figure out how to escrow money without holding it myself, and I kind of went down that rabbit hole; I still practice law, but have clients who are trying to build blockchain technology that is legally compliant. I love what Vinay and Rob and the Mattereum team are doing, the Internet of Agreements, the idea of actually baking in… I think the term, I tweeted it, what Vinay said was, “You do you, wherever you are, and the software will figure out regulatory compliance,” I think that’s very, very cool. I’m happy to talk about legal developments in the US too, if folks have questions.
Clive: Morning everybody. I’m Clive Freedman, I’m a barrister, an arbitrator and a mediator, in the chambers called 3 Verulam Buildings. I’ve been interested in anything to do with IT for many, many years. More recently, I’ve been very interested in blockchains, and in particular how one resolves disputes that arise because people are using blockchains and smart contracts, and the kind of arbitration systems one might need in order to help people resolve those disputes.
Tju: Hi, I’m Tju Liang Chua, I’m a corporate transactions lawyer from Singapore. I’ve spent most of my career working in frontier and developing jurisdictions, which, for anyone who has spent time there will know, is a trustless space, and so the excitement of really being involved in blockchain and in the crypto environment as a solution to implement what, as a practicing lawyer, the issues I’ve been facing over my career has been exciting. I’ve spent the last year working in this space, working on blockchain implementation projects as well as ICOs, and it’s a good time to be in the space.
Adam: Hello, I’m Adam Sanitt, I’m a disputes lawyer at North Rose Fulbright. My background is a banking lawyer and a disputes lawyer. I’m also part of the fintech practice at Norton Rose Fulbright, that’s been going for about five years and it’s evolved through that time, from cryptocurrency through to blockchain, distributed ledger, and that’s a very active group now at Norton Rose Fulbright. Personally, to a degree, apart from my practice as a lawyer, I also research and publish in other areas, including mathematics, medical statistics, open source system code and C++, so I try and provide a slightly different perspective to the group there.
Chris: Thank you very much. I’m Christopher Wray, Chief Legal Officer at Mattereum, and everything we talk about today is what I do, so I won’t say anything more right now.
The conference topic is blockchains and world trade. What is trade? What is business? I think a way to frame this is that trade is the negotiation of reciprocal promises that are enforceable. Obviously any lawyer is going to recognise that I’m giving a sort of coded legal analysis here in terms of the law of contract. This is a legal panel, and I think this is a very helpful and, to the general public, not an intuitive way to look at what trade is, what business is.
First, negotiation. It can be in some kind of public forum or marketplace. Originally, it would be local technology enables much more remote and perhaps larger scale marketplaces that may be more complex. Secondly, reciprocal promises. What do I mean by that? I just mean that I’ve traded something where I do something for you and you do something for me. I promised to deliver to you a tonne of wheat or to translate your documents, and you promised to give me $500. Thirdly, these promises are enforceable, which means if I don’t deliver your tonne of wheat or I don’t competently translate your document by the deadline that we agreed, then you’re able to force me to compensate you for what you’ve lost, as a result of that broken promise.
So, businesses are negotiated reciprocal promises that are enforceable. This is what trade is, and world trade just happens on a larger scale and across borders. As I said, technologies enable that: communications technology, transportation technology, legal technology, something like the idea of using arbitration. Now the parties can agree that a specific individual with a certain kind of expertise will solve a dispute they may have and decide who should be compensated what, in a place of their choosing, in a language of their choice, interpreted in accordance with the law that they choose to govern their promises. And now of course we have, in the last few decades, information technology that has also started to affect world trade and the ability to do these kind of deals internationally.
We’ll start with trying to put blockchain technology in this historical context. It is another form of information technology system; perhaps that’s not an adequate description, but that’s a starting point. Information technology systems have brought, in particular, advances in the automation of performing these promises. Perhaps the negotiating may well be automated, algorithmic trading systems take this to an extreme, but also the carrying out of those promises, actually performing them, especially in relation to financial transactions or settlement. Maybe less impact on the settlement side from information technology systems.
My first question for the panellists is what’s new here? What’s new about blockchain? To what extent does this change the kind of things that can go right with trade and the kind of things that can go wrong? Clive, I might ask you to have a few words on what could go right and what could go wrong.
Clive: Well, let’s begin by looking at the possibility of putting your contracts on a blockchain, so that software can be used to trigger an action when another event takes place, perhaps based on ascertaining a fact at a particular time; for example, the price of Brent Crude at 12:00 noon today. The most obvious risk of putting your contract in computer software is the fact that a judge can’t read the software and decide what you’ve agreed. There are several ways of dealing with that. One way of dealing with it would be to add comment lines to the computer software in ordinary English or some other language, but that may still be hard to understand, and there’s a risk that the comment lines may not be consistent with the code that follows them, or the comment lines may be insufficient, without further explanation.
A better way of dealing with it would usually be for a computer software expert to explain how the code operates and how it amounts to an agreement in an expert’s report, but again that’s adding an extra layer of complexity to the steps required to resolve a dispute, if a dispute should arise. The best method of dealing with it is to supplement the smart contract with an ordinary, written agreement. That can set out each party’s rights and obligations, it can state that in the event of inconsistency the natural language agreement should prevail, or the reverse, but that would probably be unusual. It can contain additional terms which would be important if a dispute arises, for example dispute resolution, whether a dispute should be decided by arbitration or expert determination, as opposed to by the judges in the courts; how the arbitrator should be selected, in order to ensure that you do choose as your arbitrator somebody who understand the technology; whether the parties must try to reach a settlement by mediation first and the governing law that should apply; which court system should have jurisdiction in the event of disputes generally or disputes about an arbitration. That’s how you can supplement the software that contains the contract with terms & conditions that can be enforced in a court of law or an arbitration, should that be necessary.
Let’s look at the problems that might occur, taking the example I used at the outset, of action B triggered by event A or fact A. A number of problems could occur, because things don’t happen how they’re expected to happen. First of all, the computer, which is supposed to report that event A has occurred might fail to do so. The oracle or factual source which is used to determine fact A might be unavailable at the time that it’s needed. Action B might not take place. Using computers without blockchains, the computer that is supposed to perform action B might be out of action or have no Internet connection. But with blockchains you’ve got the advantage that there isn’t a single computer, but you’ve got multiple computers in multiple locations, so that reduces the risk that the computer performing event B won’t do its job. Using smart contracts, if action B doesn’t take place, it’s likely to be because of a software error, and that of course is again a source of possible disputes.
The fourth situation is that, for whatever reason, action B might be performed when it should not have been performed. Or, fifthly, a fact might be recorded on a blockchain which should not have been recorded on a blockchain; that point perhaps applies particularly to blockchains which are being used more as a ledger than for smart contracts or cryptocurrencies. But when I was making this list of things that could lead to disputes, I had in mind initially that these might be disputes specifically related to the use of blockchains, or smart contracts as we call them, but then I found myself thinking, “Actually, they could all occur without blockchains.” The first simple example is a margin call made by a commodities broker, that’s been happening for decades. More recently obviously computers have been used to do it, but it’s all been happening day to day, without blockchains or smart contracts.
Another example which I’ve been looking at recently is online platforms for selling customer invoices to investors. Nowadays, if you’re a retailer that has lots of invoices that maybe are being paid in 30 days, you can upload the details of those invoices to a platform, and investors can buy small proportions of those invoices to spread their risk. All that is taking place without blockchains, or what we call smart contracts.
In these contexts, at any rate, where English law applies and English lawyers have got involved, there’s likely to be a 20-30-page English language contract which the lawyers have drafted, which hopefully provides the answer to the problem without there being a dispute, or,, at the very least, if there is a dispute, it tells you how the dispute should be resolved.
I think the point that could be made from these examples is that if blockchains start to be used widely in international trade, we may see disputes arising from the complicated examples that I’ve given happening more frequently, and there’s a danger that people may be over-reliant on the smart contracts which trigger the events, without spending enough time on the accompanying natural language contracts.
Chris: Thank you. Any other feelings about what might be new here? In sense, we have here that this is something that is very much staying the same.
Steve: What I would say is everything is different, nothing is different. The idea of an automated contract, if you think about the notion of a letter of credit, that is something that’s been around for a long time: if an external event happens, you have an unconditional right to pay. Now, the problem of course with a letter of credit is bank solvency. I had a case in 2008, and we settled a case in a very creative way, and it required funding a settlement with a letter of credit, and the plaintiff’s lawyer in that case, he must have known someone or something, he knew the jockey, and he insisted that we… the only bank that he would allow us to use was… I want to say it was JP Morgan. So, there’s a problem,, and I think one of the things that blockchain theoretically can solve is potentially credit risk. The problem that you face though, if you rely on blockchain to solve a credit risk problem, is you still have to have an asset, and you have to believe that asset is going to be worth something in the future.
So to my mind, what’s most interesting to me as a disputes lawyer about blockchain, and what got me interested in this space, as you mentioned, is the notion of guaranteed enforceability. I’ve gotten judgements against defendants… I had a client who had a lease with someone, and the person didn’t pay the lease: we sued them, pierced the corporate veil, got a personal judgement against this guy and against this company, but there was no way to collect it; you could get a judgement, without being able to collect it. Theoretically, a blockchain technology can perhaps, if not guarantee enforcement, provide a better enforceability mechanism. That’s one of the things that you mentioned in your introduction, that one of the components of blockchain that’s interesting from a contract standpoint is guaranteed enforceability.
If you think about contract law itself, if you go back a couple of thousand years, contract law itself was I would say an important if not technological innovation but an important knowledge innovation. It was the idea that instead of relying on self-help to resolve disputes you could rely upon state power. I guess the question is will blockchain be able to take contractual relationships and move enforcement away from state power to something else? I don’t think we will get entirely away from state power. I think there are certain areas of dispute that we’re going to have to default to court systems or to arbitrators.
I had someone call me, an insurance broker. I’ve started to get… I can tell that this technology means something, I have 10 presentations in the next two months that were scheduled for me while I was in London, my marketing person just started setting things up. Defence contractors, insurance brokers, I get calls from people, saying… An insurance broker two days ago, “My client, big energy, gas & oil company, they’re doing something blockchain related. They’re going to close soon on this deal. Does insurance cover that? What are the risks?” so I had to explain to her… She had heard a presentation that I had given, and I had to explain to her what a fork was. How do you deal with fork risk? What happens if you’ve got a smart contract that is on a blockchain and it’s forked multiple times, and that contract can trigger payment from an external bank account? How do you hedge against that?
How much money is frozen in Parity, was it $150 million in crypto? Maybe that all gets sorted out, but let’s say you’ve got a billion dollars that’s tied up because of a software glitch, because someone who has commit access to a repository they shouldn’t have changes the code. It’s only a matter of time before that ends up in front of a judge. So I think we probably don’t get completely away from state power enforcing contracts, but we move in that direction.
Adam: I agree with both of those comments, and there’s a qualified optimism there. I’d like to highlight a point, if you like, underneath the smart contract layer, which is the ability to have trustless transactions using blockchain. The paradigm for trade is that you are standing in front of someone, and you give them the money and they give you the goods at the same time. That is a trustless transaction that occurs in physical reality, because we can physically transfer these things, independently of any record of that taking place. The problem with international trade is that you are necessarily divorced in geography, in time, and we can assume because of that there is no trust there. You are trying to move sometimes information but often physical items of value over time, and the techniques that we have in international trade at the moment try to replicate that trustless situation by using trusted intermediaries, letters of credit for instance or bills of lading, to try and replicate the physical aspect of handing something over. What you’re able to do of course with blockchain is to have the analogue of that standing in front of someone and swapping something, but you’re creating a digital record in a trustless way. On top of that you can build smart contracts.
Particularly for international trade, if what the record is is itself there, exists in the blockchain, you’ve got a complete analogue or physical reality, you’ve got that automatic enforcement. But often what you’re looking at is a record of something that physically exists, and then you have all the issues that Clive pointed out. I think of them collectively as reality mismatch errors, that the blockchain is absolutely correct in the record that it’s created; unfortunately, it’s just reality that is at fault, and the two have, in some way, to be reconciled. And I think sometimes it’s not clear to people, so it’s important to stress, that there is in fact no way to avoid these reality mismatch errors. The underlying physical goods, maybe they catch fire and disappear. If the underlying physical good is a diamond, maybe someone cuts the diamond in half so you’ve now got two diamonds. Physical circumstances can change.
The other sort of reality mismatch is a problem with the event that you’ve created. You’ve said that I’m transferring this, that record exists, but perhaps that transfer is induced by a fraudulent misrepresentation, perhaps someone is pointing a gun at your head and saying, “Create this record.” In those cases, courts will reverse that transfer; in physical reality, in blockchain reality, they will do that.
So although the trustless transaction is going to help with automatic enforcement and resolve a lot of disputes – and save a huge amount of costs, as Vinay talked about initially – you are going to have these reality mismatch errors, and you are going to need the dispute resolution that Clive talked about. I’ll finish with, and I don’t know if it’s a slightly fanciful analogy, but I think of the invention of writing, several thousand years ago people saying, “Great, finally we can write down our agreements. They’ll be set out there, there will be no problem, they’ll be enforced automatically, and we won’t even need lawyers anymore, because everything will just be written down.” It didn’t quite work out like that, but it certainly did help; I think the same with blockchain.
Chris: Vinay is reminding me that the reality mismatch error is what we often talk about as delamination, this coming apart of the representation in software and the reality. So, we’ve heard a couple of times, from Clive and from others, about this need for some kind of record of the contractual agreement, of the terms of an agreement, in addition to a piece of software that’s implementing it. I don’t think there’s any voice on the panel saying that you might be able to do away with that completely.
Steve: You could.
Chris: Okay, so let’s talk a bit about this. This is the idea that a lot of our work is based on, this concept of Ricardian contracts that Ian Grigg developed originally, our Chief Scientist. You will have a piece of software which is representing value and transfers of value, but that is going to be accompanied by a natural language, a written set of terms that describe the agreement between the parties, the agreement that can actually be enforced, the intentions that they’re bound by.
Let’s think about, in the context of blockchain, what this might start to look like. Yes, we want to do business, we want to do trade with one another, we’re making promises to one another, we have written agreements that lay out what it is that we’re promising to do, and now we have perhaps smart contracts on blockchains or other pieces of blockchain technology, distributed ledger technology, that are automating part of the performance of these contracts. Perhaps also automating the enforcement of these contracts, or at least automating performance in such a way that once might would have to have been enforced, and now it doesn’t, because it has been automated.
And yet, on the other hand, you can have both automated actions, or automated states of affairs which do not involve a required action, where the automation is generating a need for enforcement. Something has happened in an automated way, or has not happened, in accordance with a piece of software, and that is not in accordance with the intentions of the people who’ve made that deal, and now we have this issue of how to enforce the deal if the software hasn’t automated it.
Steve: A couple of years ago you would see… a frequent meme was code eats law, code is law, code will get rid of lawyers. It’s easy to dismiss that, and I think I will probably be the last buggy whip maker, I’ll be around, but it’s an interesting thought experiment. Contracts have not existed for… If you go back 100,000 years ago, we didn’t have contracts, maybe 10,000 years ago… I don’t know when the first contract was, maybe the Phoenicians created the first contract… Contracts are an artefact of trade, it’s a way to memorialise what counterparties have agreed to do. But can we imagine a future in which we don’t need contracts? That would be a future in which you were able to solve an assent, agreement, you were able to create a system that removed any possibility of disagreement about what parties intend to do. The problem there is that sometimes you actually want ambiguity in an agreement, but let’s assume hypothetically there are certain agreements in which you don’t want ambiguity.
If you can imagine a scenario in which intent and assent, or agreement, can be perfectly captured, and which performance can be, for all practical purposes, guaranteed, perhaps you don’t need contracts, and perhaps you don’t need people like us to go into court and to argue with a judge. I would say that remains hypothetical, but perhaps that’s the… If you look at the horizon of blockchain technology and you look at… I don’t actually like the term “smart contracts”, because, as a thousand other people have said, they’re not either. But maybe in the future there can be such a thing, and that thing could actually get rid of the need for a contract. Maybe we won’t need to have something called a Ricardian contract, maybe we won’t need contract law. Fanciful, but possible.
Clive: Fanciful. [laughter] Let’s take the example of two cavemen. One of the cavemen has got more grain than he needs, the other caveman has got more animal skins than he needs. They meet in a cave, one hands over some grain and the other hands over some animal skins. You may say that’s not a contract, but it is a contract. Because both of them are making an implicit promise that what they’re handing over is worth having and that it is in satisfactory condition. Let’s suppose that the grain is all in a sack, and although when he puts his hand in and he takes a sample out, that looks okay, and he gives it a sniff, at the bottom of the sack there’s a whole lot of not grain. You’ve got a dispute.
Chris: But not a Ricardian contract.
Steve: If I may, there’s a difference between an agreement and a contract and a dispute. A contract presupposes enforceability by a third party. The elements of a contract in the United States are offer, acceptance, mutuality of consideration, and something that can be enforced. In that situation with the cavemen, they enforce it by hitting each other over the head with a stick, and the innovation that law provided and that courts provided was we don’t want people to hit each other over the head with sticks; we want to have Solomon decide, “Okay, split the baby,” you have a neutral arbiter decide.
Now, the question is, and I will readily admit that it’s a thought experiment, let’s fast-forward 100-200 years from now. Is it possible to imagine… Just like those cavemen could not imagine being able to go to a court and have a judge decide something for them, relying on a body of cases and case law, can we imagine what a system would look like where we’re beyond enforceability, where we don’t actually have to have a person decide that the agreement was enforceable? I think it is fanciful right now, I don’t think we have any worries about job security, but I wonder what that would look like.
Chris: Adam, to your point earlier, I think the key parts here, in the Ricardian contract concept, are yes, some software that automates something. But the other part is that, the invention of writing, it’s the clear written record of what it is that was agreed. So while we can most definitely have contracts without a written record of what was agreed… It may be unwise, because it may be rather difficult to persuade this state power that’s going to enforce something against you that that was indeed what the other person agreed to give you. So the question here I think is, more precisely, if our automation gets good enough, would we be willing to do away with a clear written record of what we’d agreed to do?
Tju: At some point you’d still need enforcement, and that does come down to power. When we evolved beyond the two cavemen hitting each other with sticks, what happened then was that they went to the village chief who then had the entire village’s worth of clubs. At the end of the day, there needs to be a resource, some form of state power, the centralised group power, whatever you call it, that is able to enforce the promises. So while I do see that we may move away from the static environment of a contract that goes directly to a court, the question of a dispute resolution mechanism will always need to be there, and it can be pre-agreed, we can pull up structures, we can move it away from the jurisdictions we’re talking about.
One of the things I do… And we’re in the Stone Age here, we do this by hand, but if you do joint venture contracts in developing jurisdictions, one of the big questions of cooperating with Chinese companies, that comes up every time, is am I going to be able to enforce my IP? I can get a judgement, I can get arbitral law and I can bring it to China. Will I be able to enforce it? One of the things blockchain does is change the mindset around it. Right now we do this by hand, we design the relationship in such a way that the reference to state power isn’t in China; you put the IP and the joint venture entity in Hong Kong or Singapore where you can rely on the courts. Restructuring the entire arrangement so that you do not rely on unreliable enforcement, state power, in a jurisdiction but move it somewhere else.
Or, for that matter, the question of designing smart contracts and matching them to law, my view is that you have really got to do those two in conjunction. The programmer will be able to develop a piece of software that automates things. Who tells the programmer how the program ought to be designed? What scenario, what treatment should the contract flow take, if event A occurs? That’s still a design that’s being advised by the lawyers drafting the contract. You can convert the natural language contract into code. I’d say that law is code, rather than code is law. In order to properly draft a contract, you do need to have a mindset of the flowchart of event A occurring leads to consequence B which leads to consequence C, which leads to a form of these possible events… That flow of work and thought process, and the understanding of the market environment and the legal conditions in each of the places where you’re operating is necessary in order to develop the software in the first place.
Chris: This is another interesting angle, and matches our own experience in initial work with clients, that the legal input in some ways is about design, it’s designing the relationship and the space of possibility in which events might unfold, what constraints do we wish to place on how things may play out. What the automation that you can have with smart contracts adds to it is that it’s not simply about saying, “Well, here’s our preferred path forward. You do exactly what you promised, I get exactly what you promised to pay me – great,” but we’re also designing in non-preferred paths through the system. It’s starting to scenario plan and figure out various things that may happen, and to some extent those can then be considered in advance rather than as a surprise, and to some extent the appropriate response is if something isn’t quite right can be programmed in, can be designed in.
It’s something that ought to have always happened. Lawyers always should have been, but probably didn’t have the opportunity to – and I won’t explore where the incentives might lie – but didn’t have the opportunity to design business arrangements in such a way that a dispute was less likely to arise in the first place, because something had been anticipated and provided for in the contract.
Steve: One of the reasons I’m not too worried about job security is I think you scratch a law firm now and you’ll find they’ve got an innovation lab or somebody who’s talking about lawyers learning how to code. As long as we exist in jurisdictions where there are laws, there will be a need for lawyers to provide input, and ultimately the lawyers are going to learn how to program, so a future lawyer… You might not have people who go to court – which is sad, because going to court is fun, juries are awesome… You don’t have civil juries in England, and they probably doesn’t make any sense, but it’s great fun – but I think our jobs will morph, and there’s something about law and design that’s very, very interesting, I think we’ll see more and more of that. Maybe eventually where it’s like Hermann Hesse and it’s like the Glass Bead Game and we move beyond law and to some sort of existence where we play complex mathematical games all day… I don’t see that happening anytime soon. So as long as we have law, we’ll definitely have to have designers who have legal training, or lawyers who have design training.
Tju: Steve, I’ve never been to court so unfortunately it’s not an exciting life that I lead. But to your point, contract drafting today is exactly like programming, it’s an iterative process. Each time Steve thinks up some new loophole that the contract does not cover, the next round we negotiate the contract has a new clause that matches that. We’re bug fixing, and I don’t see that the shift to coding is going to change that work away from lawyers.
Steve: It changes what we do. There’s someone here from Etherisc… I don’t know if folks know about this company, but what Etherisc has done is it’s one of the very few production blockchain applications that does something, and I think it’s basically Ricardian in nature. What Etherisc does is it guarantees… It’s a flight insurance application, single trigger. You go to a website, you pay, and if your flight is late, you get paid automatically. I’m an insurance coverage lawyer, I sue insurance companies for a living, and often that’s because insurance companies, it’s shocking, they don’t pay. Where is Lloyd’s from here, in which direction? That way? Hello, Lloyds! It’s Palley, I’m here! [laughter] What that does is that gets rid of certain kinds of disputes, but probably a) you need to have lawyers provide input into the design, into the construction.
Also, what’s interesting about Etherisc is they actually have an insurance policy, so it’s Ricardian in that there is actually an insurance policy written by a Maltese insurer, it’s some sort of actual insurance company, so you’ve got the paper component and then you’ve got the payment automation piece. What that will do, the trajectory there is you need to have lawyers involved in… You’ll still need folks involved in creating the contracts, but you end up moving certain things away from disputes. It’s very interesting, and I think you’ll see hopefully more applications moving in that direction.
Chris: We’ve had this code is law, but I think we’re missing something there. Code is just code; law is systems design. There isn’t just one practice of software engineering; there are different scales or levels at which you can carry out that kind of task. So I don’t think lawyers of the future will necessarily be coding, but they might well be designing the architecture of systems which some people do indeed then go and write up in code.
Adam: I think we’ve maybe morphed slightly from the future of law to the future of lawyers, which is of great importance to everyone on the panel and we could talk about for a long time. [laughter] To your point, you said it’s important to view that hypothetical of the code and just the code, whether it’s the lawyers doing it or not. I think where it may become important is for a lot of contracts and a lot of international trade are in fact right now not enforceable, in the sense that it is simply not worth it for the people doing the trade to enforce the contract. If you’re buying something worth £100 from China, you will very quickly find out the costs of trying to complain and obtain legal recourse – it is simply not worthwhile.
So there’s a huge mass of un-lawyered, effectively unenforceable contracts in the international trade space, and where we talk about this code creeping up and trying to cover the most common situations, that may well be where it enables international trade to flourish and slowly moves its way up the value chain. Yes, the larger disputes will remain the same, but I think it will still be very important. But I think that meme that we don’t need the law, even going up the value chain, I’m not sure that that is dead. We think it is maybe as lawyers up here, but I’d be interested to hear contrary views, if there are any, in the audience.
Comment: It seems to me that the other key aspect to any of this working is something that I think is inescapable, which is consent. I don’t think you can ever have anything that’s truly trustless, because the trust simply moves, because both parties still have to consent to take part in and be subject to the contract, whether it happens digitally or not. At that point trust becomes “I trust this person, because they’ve consented to agree to these terms,” and I think you can never remove that, and lack of trust means they don’t want to play by these rules, so you’re going to be wary of doing business with them, if they seem to be unwilling to play the game, to participate. So at that point trust simply becomes generated by the consent and the willingness to engage in and participate in the contract, whether it happens in meat space or whether it happens in digital.
Chris: Let’s not drop this point about how much of trade today is not practically enforceable, and how much trade is potentially not happening, because if you were to try and do a deal of that nature, it would not be practically enforceable.
Comment: Loved Tju’s comment about clubs and the pooling of clubs, the village elder. My take on it is humans have two mechanisms of enforcement: violence and discretion. Legal systems make use of violence; it’s a pooled violence, it’s according to processes, etc., but it’s still violence. We also use discretion. One of the wonderful demonstrations of a discretion-based system working in contexts where the legal system does not operate, for the very reason that just got brought up, is eBay and ratings systems. These are systems that have enforcement that has to do with your opportunity to interact with future parties, and whether that opportunity gets altered based on how you show up this time. I think that if there is a threat to law, it’s the rise of a more functional discretion-based system. It doesn’t necessarily mean law goes away entirely, but it may shift significantly.
Chris: Let me just speak to that first, because I think that’s a very important point. I do think we need to draw a distinction here between the consumer level, retail, and business-to-business, like trade at slightly higher values. For one, we lose the possibility of using escrow when we go to world trade. There’s always going to be that opportunity to tie some…
Steve: Why do you say that?
Chris: Because I don’t see that if you’re running supply chains where liquidity is at a premium that you can afford to tie up significant chunks of cash. Quite the opposite. I would say in this context enforceable legal contracts are a form of trade finance, they precisely allow you to avoid to tie up a big chunk of operating capital in escrow.
Tju: I kind of agree with Steve there. I’m actually working with a team that’s looking to build a blockchain trade finance platform for Africa, and that’s a space where the current estimates are hundreds of billions of unmet trade finance needs, because the legal systems are not… It’s a systems design question. The legal systems are such that enforcement is pretty much impossible on the ground in Africa; you have security over the asset, but there’s no way to enforce that security. There are ways to design a system where you are taking escrow, you’re taking say a crypto token escrow in lieu of a letter of credit, reducing transaction costs, structuring the… As you mentioned, systems design: thinking through how the platform works, how the incentives run, such that it does become legally enforceable in a jurisdiction where you can rely on state power.
In terms of kind of a B2B and B2C differential, I’m not sure that the reputation scores on eBay are any different from credit ratings.
Steve: The word “trustless” has always bothered me, actually. I think I came up with this, that trust didn’t go away, it just moved. What you’re doing fundamentally is you’re moving trust to Vitalik Buterin and to a four-year-old experimental blockchain. So it’s not that these systems are… I think what it means is you’re moving the trust between counterparties to trust brilliant programmers, in a system that doesn’t really have an established governance. You’re taking one set of problems and potentially creating another one. I’m very bullish on blockchain incidentally, it’s not a criticism; it’s just a practical reality of where we are.
Tju: Steve, I would say the purpose of trustless is to move the trust to someone who doesn’t have an incentive in the system. If we’re doing a trade together, if I need to trust you, you have an embedded interest in the system; but if you trust a judge, then he’s got no particular stake one way or another. That is the extent of moving things away from the need to trust counterparties in the system and move it to a reliable source, whether that reliable source is the group at large so that there’s a consensus algorithm, or whether, in current times, we use state power and appeal to courts.
Steve: I mean, I kind of like court more than I like the EIP situation right now. My point is it’s not trustless. You’re trusting people who you don’t know, they have a financial interest in the value of a cryptocurrency, and you’re also trusting miners… What I’m trying to explain to clients who are going into these things is you’re taking one set of trust problems and potentially creating another. Which isn’t a bad thing, it’s just a question of looking at it with both eyes open.
Chris: To put in a word for the legal tradition, I think the fact that you have, at least in some jurisdictions, those relatively neutral third-party institutions, the courts, that can be relied upon is a major service.
Adam: Trust has come up a few times, and trustlessness, and I sort of put it front and centre at the start. Obviously you’re moving the trust, you’re moving it to a sort of collective endeavour, and it depends on which blockchain exactly, but I think there is still more to it and there is still a genuine trustlessness that you created. We could debate exactly how that worked or exactly where you’ve moved to, maybe that’s a bit technical.
But to pick up just on the reputation side, the reputation remedy, if you like, is complementary to the trustless nature of blockchain. You don’t need blockchain to have a reputation system, and a reputation system can replace that, if you can’t achieve trustlessness. Perhaps the analogue for the reputation type remedies within the blockchain system is the ability to exclude bad players as a punishment for breaking the rules, so a type of enforcement which can be done without recourse to the legal system, where you say you are no longer allowed to transact on this blockchain, you are excluded from it in some way. That can be very powerful, that can be done internally to the blockchain. It raises lots of other issues when you’re dealing with multilateral as opposed to bilateral situations. You’ve now got more than two people involved, you have the whole system involved. How do you obtain consent to that, how does that work? I think those sorts of remedies will be important in future blockchain systems.
Chris: Would you prefer a neutral independent party, or would you prefer a popularity contest? We did have a question at the back there.
Question: What confuses me in terms of this contract is the enforceability. I understand binary, one-off payments, options, futures, all of that, but how does it work in a more complex contract? You mentioned IP rights and enforceability. What’s the point of me having an IP right, if I can’t within the code of the contract ensure that the PRC will enforce in Shanghai my rights. It’s that break with the reality that I… Can you please speak to that?
Chris: Yes, thank you – I think that’s one of the most important points we can make this afternoon. Yes, there will be things that we can fully automate. If I’m going to sell you some Ether and you’re going to send me some Bitcoin, those are digital assets and the transactions can happen immediately. If I’m supposed to have the right to receive some oil physically from some location for a period of time, it doesn’t really matter what any software system says, if I stop receiving my oil from that location.
Steve: The distinction I make is between single trigger payment, where you’re relying on a trusted data source – I don’t like “oracle”, because the word means the opposite of what people intend – that’s one thing. But the question is how do you automate performance using a smart contract?
I think actually some fairly complex automation is possible. Do people know what a quantity take-off is in construction? Quantity take-off is when you’re doing excavation, you’ve got to pay for the cost of the amount of dirt that you remove, and one of the most common disputes in American construction projects is the amount of the quantity take-off. How do you figure that out? Do you weigh the soil? Do you eyeball it? How would you automate that? I thought about this, I did a presentation for the ABA three years ago, we brought a drone in, nobody showed up. I actually talked about Ethereum and I spelled it wrong, I’m still kind of embarrassed, in my white paper. What we discovered was you can actually automate something that seems very performance related, quantity take-off, and you would use a drone that uses photogrammetry to look at the earth, measure the amount of soil that’s removed – you can do that actually more precisely using technology than you can using a scale, which has to deal with the variable weight of a truck – the drone can then send that information to a smart contract which can then automate payment.
The problem with trying to drain the ocean at once is it becomes impossible; you have to think about it from an iterative standpoint. So what I would do in your situation is take the most complex performance issue that you have that’s something that you want to try and automate, speak with programmers, and basically you break it into the smallest possible pieces. With construction, which is one of the things that I focus on, I would begin with excavation, quantity take-off, and then you look at things that are more complicated. How do you figure out if the millwork was installed per plan, per spec and done on time? Can you automate that? I think what you find is if you break things down into their component pieces, you can actually begin to automate fairly complex performance requirements as well.
Chris: My follow-on to that, and then we’ll come to your question, is consequential damages, consequential loss. This is the thing: it’s one thing to have a contract where we can automate enforcement, we can take something that would have perhaps required enforcement action, you’ve delivered short, and we’ll reduce the payment accordingly. That no longer needs to be enforced, because that’s now a non-preferred pathway system that can be automated. But, if you knew full well that I needed that delivery from you because a whole series of other things depended on it, it’s going to be potentially what I owe you to make good for me breaking my promise is more than you’re ever going to pay me, had I delivered. In these kind of contexts, it’s hard to see how any kind of escrowing solution can work. I think that’s an interesting question, what the economy looked like, when nobody can get consequential loss.
Vinay: This seems like this is where we get into this very fuzzy middle ground, where reputation, insurance and escrow all begin to interconnect to each other: network escrow, trade associations, various ways of doing long-term accountability. It seems like we’ve seen over many different technological substrates many different ways of approaching those problems, and David Friedman’s The Machinery of Freedom illustrates a whole bunch of examples. I tend to think that reputation systems are very fragile, I’d much prefer things like insurance and escrow. But is there any work, does anybody know of any work that looks on these things on a spectrum, on a network, that has some kind of nomenclature and analysis of how we move systems from an escrow norm to an insurance norm for example? There must be academic work on it, I just don’t know where you would find it. That might be a question for the audience as much as the panel.
Steve: Escrow and insurance are bonding in the United States, they’re different things. Escrow is a payment guarantee, it’s a two-party payment guarantee; insurance is contingent risk, you pay a premium to a third party, so you’re pulling the asset. I don’t know that you would necessarily ever move entirely from escrow to insurance, or from insurance to escrow, I think they serve… If somebody has a different opinion, maybe I’m wrong, but I think fundamentally they’re different relationships, one is tri-party and the other is probably two-party, and I think they serve different risk functions, but I don’t know.
Tju: Well, certainly from the buyer standpoint they serve one and the same thing, which is “I want to get paid.” Whether it’s through escrow, whether it’s from insurance, whether it’s from offloading the asset to a third party market, I just want to get paid. So as long as the system is designed correctly for the result that I need, the mechanism, as a participant in the system, is, frankly, irrelevant. In the M&A context for instance, there’s been an increasing market for insurance over breach of reps and warranties. You’d think that that’d be the most absurd thing in the world, but there’s a market for it, because it’s priceable; where it’s priceable, there’s a deal to be done.
Question: I’m not a lawyer, so excuse me for maybe a slightly basic question, but I get extremely upset when people talk about trustless trust, without actually defining what it means. People like Chris Dixon of Andreessen Horowitz recently published an hour-long podcast where he mentioned trustless trust at least 50 times, saying it’s going to change everything, without actually specifying what it does change. As an ex-banker, when I look at trustless trust, these types of contracts, agreements simply look to me as fully collateralised agreements. In the context of things like B2B, or essentially trade financing, that’s why banks came into existence, to provide trust into contracts, where otherwise you would need huge amounts of collateral. Could you in a way help clarify, from your perspective, does trustless trust actually mean anything? Specifically, does it lead to disintermediation of the framework otherwise surrounding these contracts?
Chris: Excellent question, thank you.
Adam: We brought up trustlessness at the start, and it is important, and we’re not in a sort of nihilistic worldview where there is no trust in anything. To go back to the first example of swapping the goods for the money, there is a trust there, there is a trust in the fiat currency that you’re handing over, a trust that that system exists and that that currency has value. I think you can sort of go down the rabbit hole trying to locate where this trust has moved to. It has in some sense moved to a collective system, but I’m slightly reluctant to go more into that, just because you can spend hours looking at podcasts. I wonder if we could stipulate here that the trust has moved so that, as you said, you are able to disintermediate, without getting into a sort of theological debate as to the nature of that trustlessness. Would that help?
Comment: Picking up on that point, I have a slightly challenging view for you guys, which is that actually as a lawyer you learn eventually or quickly that contracts are really about relationships, and most contracts aren’t actually enforced, and they’re certainly not enforced in accordance with their terms. There’s a lot of anthropological evidence that suggests that societies have prospered the most where there’s not just been the ability to trade but there’s been trust, and I think this idea of trustlessness as a goal, beyond payment and cryptocurrencies, is not necessarily a good thing.
The legal systems which have proved most useful for trade are Anglo-Saxon legal systems, not civil law legal systems, and Anglo-Saxon legal systems have always baked in discretion. You aren’t held to your promises. That’s why you don’t have specific performance, you have damages as a remedy, that’s why there are equitable remedies. The fact that you won’t necessarily be held to your performance is actually part of what builds the trust in the relationship, people have to be almost able to not perform or decide not to perform, and that builds the trust. So I’m not convinced that using blockchain technology is necessarily a good on its own just because it enables trustlessness. I think we want to encourage trust between people, not make them feel that the world would be better if they could trade trustlessly.
Chris: I don’t disagree with you. I think one of the nice applications of blockchain technology, in a Ricardian contract context, is supporting this. Trust but verify might be the slogan. If you have scenario planned, if you’ve actually done a bit of design about the kind of relationships you want, and you have an idea about what’s going to happen and when, and what kinds of evidence there would be of performance as this contractual performance unfolds, you can take those pieces of evidence, you can hash them, you can put the hashes on a blockchain. That’s a really nice way of ensuring that everyone knows that if there were to be a dispute, any neutral party coming in to resolve it is going to have a nice chronology, and it’s going to be very difficult to resile from those notices that were sent, and you don’t even need a…
Comment: It’s not about disputes though, is it? It’s about relationships. The world changes, people decide to do things in different ways, and a contract is kind of a background to how you have that discussion or how you have that relationship. In the real world it’s not a contract that gets enforced; it’s more “What’s the reason? How can I help you? What we can do about that?” That’s how business really works.
Tju: I would have to disagree with that actually, as a systems design question. Because trust has a scaling problem. You can trust the people you know, you trust the immediate people you trade with, you can discover whether or not you can trust a long-term partner. But when you buy a random item that you want off eBay, you’re trusting eBay because you have the relationship; you’re not trusting the seller, because you don’t know the seller. So a system that allows the shifting of trust to a mechanism that you can build a one-on-one connection with enables global trade, without which we’d all just be trading with our five nearest neighbours.
Steve: The same thing with insurance policies. I’ve been dealing with insurance companies for 20 years, and insurance companies don’t treat everybody the same way. Insurance companies aren’t all bad, we need them in order to operate, but there is a different relationship there than you have between maybe two commercial entities who are doing business, maybe a developer and a contractor. If I go back to my example of the quantity take-off automation, I think that that can help build trust. It removes one of the areas of dispute that you can possibly have and facilitates maybe a stronger relationship, so where there are those moments where you do need discretion, where you say, “You know, I know I don’t really owe you $100,000. Maybe you take 50 and I’ll throw the next job to you?” You’re right, that’s partly how business gets done. You’ve got the contract, but if you want to maintain a mercantile relationship, sometimes discretion is the thing that allows business to proceed.
Adam: Not only are most contracts not enforced, but most contracts are never actually read. [laughter] Usually when we have a dispute, they bring out the thousand pages of contractual documentation from a drawer where they’ve been gathering dust, and no one has actually looked at them or conducted their relationship in accordance with that contract, and no one looked at it until there was actually a dispute. So yeah, there’s certainly some truth to that. I would just say I think it’s important to distinguish the ability to create a record, let’s say to know what’s in a bank account, without having to trust a third party bank to have those records, in the same way that you can transfer something to someone when you meet them face to face. It’s important to distinguish that and to recognise that advance from the trust in the underlying relationship. We’re not advocating a lack of trust there; we’re just adding some new things that can be done that seem particularly relevant for international trade, where you are separated by geography and language and time and in a different trust arena. But how it will work I think depends on the factors that you bring up.
Question: Thank you, Chris. I’m the leading member of the Vinay fan club [laughter], and Dominick, my co-director next to me, not to exclude you, his fan club too. This is really for Mr Palley. I was an American and international lawyer. I don’t practice anymore, I went into financial services and fintech some time ago. But when I was a lawyer, I wrote a professional… I’m conscious, having lived here for many years, that everything we’re talking about is in the context really of the UK or other countries. So when what we’re talking about today – Mattereum, Internet of Agreements – is resolved, that will be the situation, the law in a nation state, it will be the law of the land; not so in the States, and that’s what I’d like to ask you about. Years ago I wrote a professional law review paper, which somehow became the bleeding authority for something like 10 years on international letters of credit, and people asked me to be part of a working group and we put together something very similar to what’s being done here. But despite the uniform commercial code – you’re smiling already – the 50 states, by the time they were through, tore it to pieces and there was absolutely no uniformity of anything. So whatever this august body comes up with in time will be the law in the UK, in France, in Estonia, in wherever. But please, tell us what happens when it gets to the United States.
Steve: The United States is not unique in having a federal system, but it does create… It creates opportunities, but it’s complex. Every state has its own legislature and has control over certain bodies of law. Contract law, for the most part, in the United States is a matter of state law, and what the law is in Utah about contract law isn’t necessarily what the law is in Idaho. It’s particularly nettlesome for insurance, because of something called… There’s a statute that basically says that insurance is regulated by the state. Every single state in the United States has its own insurance commissioner, and if you want to be an insurance company in the United States, unless you’re a certain type of health insurer, you can’t write insurance for the entire country, so if you want to write insurance policies… I don’t know if folks know about Lemonade, it’s an interesting insurance company that’s launched in the US, they’ve got a very famous behavioural scientist involved in it, but they had to start in New York.
So I think you’re right, what happens in one state in the United States doesn’t necessarily happen everywhere else. Unless Congress takes something away from the state and that’s not challenged by the Supreme Court, we don’t have the same national uniformity.
Chris: The final comment should be though, with all this talk of fragmentation, let’s not forget the wonders of arbitration. This is about world trade, it is rather wonderful that you can choose your law and specify the kind of expertise you want your arbitrator to have, and have somebody competent decide what it was that you agreed and what damages who should pay to whom, based on a law that you expected to be applied by someone with the relevant skills. Arbitration is a wonderful thing.
Thank you very much to my panellists. That was, I hope you’ll agree, a very interesting session – let’s give the panellists a round of applause, please. [applause]