E03 -

Kate Sills joins to respond to several recent articles criticizing smart contracts, including the “Oracle problem.”

Paul Matzko
Tech & Innovation Editor

Kate Sills holds degrees in Computer Science and Cognitive Science from UC Berkeley

Matthew Feeney is the director of Cato’s Project on Emerging Technologies, where he works on issues concerning the intersection of new technologies and civil liberties. . Before coming to Cato, Matthew worked at Reason magazine as assistant editor of Rea​son​.com. He has also worked at The American Conservative, the Liberal Democrats, and the Institute of Economic Affairs. Matthew is a dual British/​American citizen and received both his B.A and M.A in philosophy from the University of Reading in England.

This week Kate Sills joins us to respond to several recent articles criticizing smart contracts. One of the issues we cover is the “Oracle problem,” which is how a self‐​enforcing digital contract can know whether terms have been met in the physical world. Also, we discusses how smart contracts cannot be mere transactional documents but also need to facilitate relationships, something that has led traditional contract law to purposefully include ambiguous or unenforceable clauses. Finally, we talk about Alex Tabbarok’s call for a dedicated arbitration system for smart contract disputes. Disclaimer: While squirrel mortality is referenced briefly during the episode, no members of the Sciuridae family were harmed in the making of this show.

Kate Sill’s case for smart contracts.
Jimmy Song on the Oracle problem.
Alex Tabbarok calls for a smart contract arbitration system.

(Deep Cut) Karen Levy reminds us that contracts are relational documents.
Listen to Kate Sills discuss smart contracts and the blockchain on Free Thoughts.


00:00 Paul Matsko: Welcome to Building Tomorrow. I’m your regular host, Paul Matsko. With me today in the studio…

00:06 Matthew Feeney: I’m Matthew Feeney, the Director of Cato’s project on Emerging Technologies.

00:10 Paul Matsko: Today we’re gonna be talking about smart contracts and we’ve brought on a special guest with some expertise in the area. I’d like to welcome Kate Sills to the show. She came on Free Thoughts a couple of months ago to talk about smart contracts as well from a different perspective. Kate, why don’t you tell our listeners a little bit about yourself, what you do.

00:32 Kate Sills: Sure. For the past year or so, I’ve been researching smart contracts. My background is in computer science. I’m a software developer. I just recently started at the startup, Agork, and the co‐​founders have been working on secure smart contracts for probably about 30 years now.

00:48 Paul Matsko: Congratulations.

00:50 Kate Sills: I’m so pleased to join you today, and thanks so much for inviting me.

00:54 Paul Matsko: Yeah, well, we’re excited to have you. And I think I saw on your Medium bio that you also are a Tezos Commons board member?

01:04 Kate Sills: Right, that’s correct, yeah. The Tezos Commons Foundation is a non‐​profit in support of the Tezos blockchain, which will be having a beta net very soon and then a main net launch after that, so, that’s also exciting.

01:16 Paul Matsko: That’s really cool. Well, I think we might start at ground zero here for our listeners who aren’t incredibly familiar with smart contracts. They’re probably familiar with the concept, they’ve heard it bandied about, but could you start us off Kate by explaining what a smart contract is and what’s the problem it’s devised to solve?

01:41 Kate Sills: Sure. I think this is a very contentious term, so I’m just gonna throw that out there. But my personal definition of a smart contract is a credible commitment that’s enforced in code, and I can try to break that down. But I think first off, it’s very important to note that a smart contract doesn’t necessarily have to be on a blockchain and in fact, smart contracts predated blockchains. They come from the early ‘90s and they were in use even before the World Wide Web started. So there was an entity called AMIX, the American Information Exchange, that allowed people to buy and sell information, kind of like consulting time. And that used a form of smart contracts. But the problem that smart contracts are really trying to solve is what legal scholar, Anthony Kronman, calls transactional insecurity. And this is something that goes back to Hobbes, where it’s people wanna make a deal, but once one side has carried out their part of the deal, the other side may renege on their promise. And this is what Oliver Williamson called opportunism. So that kind of transactional insecurity has always been a problem, and it’s a problem on the Internet. And so, smart contracts are trying to solve that.

03:01 Paul Matsko: Now, so traditional contracts… I suppose we should say traditional contracts instead of dumb contracts, dumb versus smart. Traditional contracts, this is where you would go to court to enforce the contract. If someone doesn’t abide by the terms of the transaction, you take them to court or threaten to take them to court to force compliance. And the idea here is that the code in the smart contract, the automatic enforcement obviates that need. You don’t have to do that anymore. You don’t have to go to court because the contract does what the contract’s going to do as long as the one party meets the terms of the transaction. Am I getting that right? Is that how the smart contract functions?

03:47 Kate Sills: Yeah, I think that’s exactly right, and I think that’s a really important distinction, because in the real world, in the brick and mortar world, we can’t just say, “Here are the rules,” and then people automatically follow them. That’s just not how people work. So we have to have this kind of exposed enforcement where if you break the rules, you get punished. But in code, in the computer world, it’s very different. You can actually just say to the computer program, “Here are the rules,” and the computer, being a dumb machine, will actually just follow those rules, presuming that you coded everything correctly. So this is something that Lawrence Lessig talked about in Code is Law is the Internet and computers in general open up this whole new world for constraints on human behavior. Where if you write something in code, it just runs exactly as you coded it, whether that’s a good thing or a bad thing. So it’s not this exposed enforcement, it’s actually enforcement just by running the code.

04:49 Matthew Feeney: And how is this better than what we might call traditional contracts? If I wanted to enter into a contract with Paul about… Well, next week I promise that I’ll sell you my car or something. Why would a smart contract be preferable to what we’ve been used to?

05:09 Kate Sills: Sure, sure. Just to say first, I think there is a lot of hype. I think it’s not necessarily true that at this point in time smart contracts are even better than legal contracts in a lot of ways. But I think people also have an unrealistic perspective of legal contracts. Right now, the majority of the world doesn’t have access to court enforcement whatsoever. It’s simply too expensive. Legal scholar, Gillian Hadfield, has this number where I think it’s under $100,000 it’s just not worth going to court. And so access to justice is a very big deal, so I think we have to make sure that we’re comparing a realistic perspective of smart contracts with the realistic perspective of legal contracts. So that being said, I think there are definitely some things that smart contracts can do better.

06:01 Kate Sills: So specifically in the area of, you have one stranger trying to make a transaction with another stranger across legal jurisdictions, and that’s something that up until now has mostly been limited to multinational firms who can afford to figure out all of the legal specificities and details that need to happen in order to actually make that transaction occur. But something as simple as a peer‐​to‐​peer auction, like an eBay model, traditionally, it’s been very hard to do. EBay has some reputation mechanisms that try to make that happen, but something like that is very, very simple to do in a smart contract. So I think smart contracts are in their infancy, they definitely need to be built out more. There are a number of limitations, but when it comes to the new economy that emerges on the Internet and that I think we’re starting to see now, once we have digital cash in cryptocurrencies, we’re going to need that kind of very low transaction cost, stranger to stranger transactional security that we haven’t seen before.

07:13 Matthew Feeney: So what is the relationship between smart contracts and cryptocurrencies? At least in some of the reading, it’s very common to read about Ethereum in particular. So what is Ethereum and what’s its relationship to smart contracts?

07:30 Kate Sills: Sure. So again, smart contracts have been around since the early ‘90s, so they predate blockchains. But what blockchains were able to provide the world was a way of solving the double spin problem. So this was a way of giving the world digital cash that they could actually transact with on the Internet in a way that didn’t require a central authority. It’s like, “Okay, well, now we have the cash. We can send money to people, but if we’re actually trying to pay someone for an item in that peer‐​to‐​peer market, we need to have some way of ensuring that I don’t send them the money and they never send me the item,” that sort of thing.

08:16 Kate Sills: What’s interesting is that the pseudo‐​anonymous founder of Bitcoin was very interested in these sorts of exchanges. There’s opcodes, so that’s the very low level commands in the Bitcoin script that are for escrow and smart contracting and other kind of more complex tasks. But Bitcoin itself doesn’t really support a lot of those, they were put in there in anticipation of the future. And so what people were looking for in Ethereum was a way to do more complex transactions, to have more clauses, more cases and just be able to handle a lot more complexity. So that’s why you would want a more Turing complete language. And so from there, there’s been this explosion of, “Oh, okay, we can do smart contracts on a blockchain and we can use cryptocurrencies as part of that transaction.” Because if we were trying to use cash, then we would have to still go back to, “Well, how are we having US dollars on the Internet? Who’s actually backing that? Who’s promising that there’s US dollars behind that?” So, what cryptocurrencies allow us to do is to include some kind of payment and transactions that we know exists and we’re not depending on a central authority for.

09:45 Matthew Feeney: So what I hear is, and to put this in layman’s terms, the smart contract is doing two kind of central things. First of all, it solves the problem of escrow, like when you talk about US dollars. Right now somebody has to hold those US dollars. A third party such that, say you’re on eBay and you wanna buy a widget. The widget costs however many dollars. There’s someone with a widget who’s willing to part for that many dollars, but the widget has to change hands non‐​instantaneously, it has to go through the mail. And so in theory, you’re supposed to get the payment and send the widget, but there’s that period of uncertainty. So eBay or PayPal or so, they hold on to the money in escrow to make sure that both parties, that person who bought the widget gets the widget and the person that sold the widget gets the money. Whether it’s Ethereum or something else, that basically crypto gets rid of that need for escrow because it’s there on the blockchain, it’s decentralized, there’s no need for centralized escrow party. And then the other bit is that self‐​enforcing, where in theory, you don’t have to rely on the courts to adjudicate the situation. So it’s like a self‐​enforcing contract and it takes care of the escrow problem. Is that a fair way of characterizing what’s going on there?

11:16 Kate Sills: Yeah. It’s definitely self‐​enforcing, and I think even a lot of smart contracts will include escrow. But the important thing is that no one is in charge of the escrow. It’s only the smart contract, so that part is self‐​enforcing as well, which is nice because then you don’t have to depend on a third party to hold your money, especially if it’s a large amount.

11:38 Matthew Feeney: Right, okay. Yeah, that makes sense. And I was thinking, in response to what you were saying about the potential benefits of a smart contract and the fact that with traditional contracts, especially folks who don’t have lots of money, who aren’t large corporations, who are not middle or upper class, don’t really access to contract disputes just because of the sheer cost of hiring a lawyer, the cost of going to court. I was thinking of rental situations. Often when I read about smart contracts, I hear folks give the example of a landlord having a lock box on their house and they do a smart contract with a potential renter who, in exchange for payment, they get access to the lock box or access to a digital lock on the door. If they don’t pay, the door locks, the door won’t open.

12:34 Matthew Feeney: But there’s some recourse there as well. Currently, if you have a rental dispute, it’s very rare if someone doesn’t pay the rent, the landlord is basically stuck with the bill for, depending on the state, for up to six months or a year. It’s often not worth it for the landlord to go through the court proceedings that are required to evict a renter. It’s just too expensive, too much of a bother, too much of a pain. So you have this whole area that goes actually under‐​litigated in a sense that, in theory, a smart contract could provide an alternative, a resolution. So I think that’s quite intriguing. What other areas do you see the potential applications for smart contracts? Like, where you see this going in the immediate future. The areas where you think that will benefit right now from smart contract adoption.

13:30 Kate Sills: Sure, sure, yeah. I think the eBay model of transactions over the Internet is a really big use case simply because most of those transactions are quite small and aren’t covered by your traditional contracting system. Let’s see, I think some other examples. So you had mentioned rental situations. I think in the future, it might be possible to actually have property title on a blockchain if the legal jurisdiction in the area agrees to that. And then from there, as long as you have a token representing the title to a property, you can use smart contracts to securely transfer those. And so, there you leave out a lot of the ambiguity that today we rely on title insurance and a whole army of people to try to figure out. Whereas, if you’re able to show, “I have ownership of this token,” it makes the situation a lot easier.

14:34 Kate Sills: There’s still a lot of problems to be worked out there in terms of, is the legal entity that’s issuing all of these tokens, have they resolved any disputes in terms of ownership before they actually issue the tokens and that sort of thing? So there’s still a lot, I think, to resolve there. Let’s see, some other examples that I’m thinking of mostly have to do with moving something that’s in the real world onto the blockchain and then enforcing secure transfer of those. But we can also see that there might be things maybe like buying and selling event tickets, things that are created property in a way, that can be bought and sold. In the contracting world, there’s a lot of different contracts. There’s marriage contracts, there’s employment contracts, there’s a lot of different things but I think what smart contracts are really great at is this kind of solving the transactional insecurity problem, especially with Internet transactions.

15:46 Matthew Feeney: So, I suppose this might not actually be a question so much about smart contracts but about the role of the legal systems we’re used to in a world where smart contracts are used. So going back to the example of, “I promise to sell Paul a car and we organize this agreement using a smart contract and then on the day of the sale, the funds get transferred and the car unlocks automatically after the contract is executed and then Paul gets into the car and it turns out that I’ve sold him a defective car. It’s not the car that was advertised and the tires are flat or the breaks don’t work, something like that.” In this world in which smart contracts are ubiquitous, I suppose, do we still have a court that Paul could walk into? Or is there something else that could potentially be used?

16:39 Kate Sills: Yeah, so I think it could be possible that we still have court enforcement of things like that but I think something also that’s really promising is the combination of smart contracts with private arbitration. So it might be that most of the contract is simply enforced in code but we leave open the possibility that if there is a dispute in that contract, that it goes to a predetermined, private arbitration agency that we’ve given the authority to decide certain things. So, the really cool thing here is that, unlike a legal contract where we’re not quite sure what the court might decide at the end of it, we can limit the discretion and the authority that’s given to private arbitration to say, “Here are the possible outcomes that we want you to decide. You can only decide among these outcomes, and then we’ll take whatever you give us, whatever answer you give us, and then put that into the smart contract.” So I think what it does is that it actually severely limits the authority that you’re handing over and allows you to have more predictability in what the total outcome of the contract is going to be.

17:56 Matthew Feeney: So I guess one possibility would be that this dispute arises and we think, “Well, we both know Kate and she’s smart and we both trust her, she can sort this out.” That’s one model but I worry about if some of these end up in court, why would a court treat these kind of agreements as anything different to an online chat? They’re just looking at a piece of computer code and trying to figure out why should that be any more important or legitimate than Paul and I just trading emails about the car?

18:33 Paul Matsko: I’m thinking about this after the embarrassment of various elderly senators trying to understand how Facebook works during the Cambridge Analytica hearings, where we’re talking about judges who are probably gonna skew older, who are gonna be more likely be tech illiterate trying to suss out the whys and wherewithals of code, right? I mean, they don’t even know what a blockchain is, let alone how a smart contract works. What it sounds like, Kate, is that this reminds me of a proposal that Alex Tabarrok made in a piece for Marginal Revolution towards international courts for smart contract arbitration, where he basically points out, “Look, there’s this long history of private arbitration between private parties that’s extrajudicial, that’s outside of the formal state court system.” He goes all the way back to, he makes a reference to lex mercatoria, the Middle Ages and merchants who have this complex arbitration system that was like court‐​like but not officially part of the legal system that arguably was a lot better than the legal system of the time and proposes a lex cryptographia.

20:01 Paul Matsko: I suppose you could build this into a smart contract system where people say, part of this smart contract is if we have a disagreement in those ambiguities. If I sold Matthew a lemon, a squirrel had died in the exhaust and now the car stank and he didn’t know that until he climbed in. Written into the contract is that the court of first appeal would be this lex cryptographia, would be this private arbitration system that would resolve some large percentage of these disagreements. You could probably write that into the contract itself. Have you heard proposals other than Tabarrok’s as people have been writing about this space, Kate?

20:50 Kate Sills: Yeah, yeah. The law merchant and using that as the framework for these cross jurisdictional dispute resolution frameworks has been something that I’ve been arguing for quite a long time. This is actually happening in the blockchain space. There are companies like Claros and I think Sagewise as well and I believe there’s another one called Juris that have noticed that smart contracts are going to need some kind of dispute resolution system and they’re trying to provide that. Some of those white papers are really, really fantastic works in, I think, law and economics and computer science. So I’m very excited about those.

21:33 Kate Sills: But I think it’s a really, really good point that this idea that we don’t necessarily need the courts for enforcement, that we can have private arbitration, it’s not a new idea. Private arbitration is used all the time in commercial settings and it goes back all the way to Medieval times where the merchant class wanted to be able to resolve their disputes very quickly. They didn’t wanna have to go to the local feudal lord or the ecclesiastical courts because those might be far away when they’re travelling for business and they probably don’t even understand what the contract is about because they’re not business savvy. So, the law merchant allowed them to choose arbitration entities that were people that were businessmen who understood where they were coming from and had a similar interest in trying to resolve disputes very quickly. So maybe it might take an hour versus months or years or weeks. I think that’s a great model for the kind thing that we need for smart contracts on the Internet.

22:40 Paul Matsko: If we’re using smart contracts on the blockchain and we use a currency like Ethereum, there’s a large amount of security that’s the responsibility of the owner of those tokens, right? What happens if I lose my key in this world?

23:05 Kate Sills: Right, right. That’s something that’s been really widely discussed because the experimental frameworks that are out there haven’t really covered this part of it. They don’t really cover the user experience and the user interface aspects really at all, which is unfortunate, but I think that’s just a matter of how early it is. I think there’s a couple of things that could be done there. One is having better software on the front end, allowing users to not just have to use the command line or have to store all of this stuff but figure it all out themselves but rather having software that assists them in that. Another is having third parties like Coinbase hold some of that secret information for them and manage it for them.

23:57 Kate Sills: But I think one thing that’s really intriguing is the possibility of having governance for the entire blockchain, and this is something that Tezos and others have been looking into, and that governance good in the future include some of that very low level blockchain‐​specific dispute resolution, where it’s not necessarily that you have a dispute in a smart contract, but that you actually need some kind of assistance on the blockchain level itself. You’ve lost your private keys. You had, in the case of property title on a blockchain, you would need to be able to go to whoever had issued that property title and say, “Hey, I’m the rightful owner still, I just don’t have the key anymore.” So that’s something that people are still working out. There’s no intrinsic problem that makes it unsolvable or intractable. It’s mostly just that people have been like, “Well, we really need to kind of like figure out the consensus mechanisms and the infrastructure for this world, and we can worry about the small user interface problems later.” And so I think that’s where we are in the building process.

25:09 Paul Matsko: It’s kind of a non‐​unique harm in a sense, where currently, I’m trying to sell my car, what’s the first thing you do? You go and dig around in your paperwork and hope that you kept the car title somewhere ’cause you don’t wanna keep it in your car and then you’re like, “Oh, crap. Last time I moved, it got put in the wrong files, now I don’t know where the car title is.” And like, “Ugh, this is a pain. I have to go see where I registered it,” and go down to some state authority and find out. This is currently a problem with losing title, losing access and you end up being thrown to… In a sense, to a third party who’s holding that information for you. So just as that is solvable, so too in theory, is this solvable? At least potentially it could be solvable in a way that’s slippier, that has less friction to it and it’s less of a pain than going to the auto registration office and hearing a bored bureaucrat give you a hard time for losing your title, and then paying to have it re‐​issued, and the like.

26:11 Paul Matsko: Something else I wanted to touch on, Kate, and for our listeners, this will make you sound really smart if you bring this up in a conversation about smart contracts at your next cocktail party. I’m referencing an article by a Jimmy Song for Medium, The Truth About Smart Contracts, and he’s very, very bearish on the potential for smart contracts, and really he’s echoing a lot of the cautions that you, yourself have issued here, which is like, “Look, there’s still stuff being worked out. This is still a work in progress.” But he mentions in the piece, he spends a great deal time talking about the Oracle problem, which is a riff off of the Oracle of Delphi, which was a real historical like Ancient Greece, and if you wanted knowledge of the future, you’re supposed to go to Oracle, pay a fee. It’s kind of like the movie, 300.

27:06 Matthew Feeney: She was ingesting volcanic fumes, or something, right?

27:10 Paul Matsko: Yes, yeah.

27:11 Matthew Feeney: It’s not a wise, old person.

27:12 Paul Matsko: Right, right, yeah. Well, and then in theory it was really, her utterings were interpreted by a body. 300’s a very loose and traumatic and titillating interpretation.

27:25 Matthew Feeney: Horrible movie.

27:25 Paul Matsko: Yeah, yes. Bad for a whole bunch of reasons. So, there you go. Sometime we can talk about technology and movies. So the Oracle problem, what is that? What is Song talking about? And I think we’ve danced around it a bit in our conversation so far, but can you flesh that out for our audience, Kate?

27:41 Kate Sills: Sure. It comes from a certain characteristic of code that runs on a blockchain, so anything that runs on a blockchain has to be deterministic, which just means that if you put in the same inputs, you need to get the same outputs every time. So if I have a function, you can just think of it as a mathematical function, and I put the number 1 in, and I get 2 out, then when I try that again, and I put 1 in again, I should also get 2 the next time. It needs to be deterministic because what a blockchain is doing is having all of these different computers, all of these different nodes running the same code, and it needs to come to the same conclusion because if the computers are all saying different things, then you don’t really have the blockchain network anymore.

28:33 Kate Sills: So code that runs on a blockchain needs to be deterministic, and so that means that a smart contract, when it’s trying to look for information, like let’s say we have a bet on the outcome of a NBA game or something like that, it can’t go to NBA​.com and pull down the info, the score of the game, because I think this has happened to all of us, we go to a link on a website and it turns out that the website is no longer there, or something has changed that it’s broken. And so code that runs on a blockchain can’t reach out into the real world and pull in information. So that is definitely a problem and a limitation, but there’s an easy solution to that, which is to have entities that write this kind of information to the blockchain so that it’s completely deterministic and the code on the blockchain can read it. And those entities are called Oracles.

29:35 Kate Sills: So the Oracle problem, and I think this is what Jimmy Song is referring to, is that you have entities that write code to the blockchain, but the blockchain is supposed to be minimizing trust, so then what are we to think of these trusted entities that are writing code or writing the results, writing information to the blockchain? So I think this is a concern. We need to make sure that whatever we’re relying on for information is actually trustworthy, but I think it’s really a mistake, and it’s really a mistake in terms of thinking of what blockchains are for. They’re not trying to get rid of trust entirely. It’s supposed to minimize trust. It’s a risk management. You can never get the risk down to zero, but you can try to limit it and limit the authority that you’re granting to other people. So there are different ways that you can try to combat this problem instead of just relying on one entity, you can rely on a lot and then average them or otherwise, combined the answers so that if one of them is corrupted somehow, you don’t have a problem.

30:57 Kate Sills: There are other things that you can do. If you think that your threat model is that someone’s going to try to bribe whoever is giving this answer, then you might just have a pool of a million people and select one of them randomly and ask them, “Hey, can you look at NBA​.com and tell me the score?” And then that way, it’s very hard to bribe someone ahead of time because the attacker doesn’t actually know who it’s going to be. And whoever selected is probably not very interested in the outcome of your smart contract. So that sort of model solves what Jimmy Song is referring to as the Oracle problem.

31:36 Paul Matsko: This is a very William F. Buckley‐​esque in a way where he said, “I would rather be governed by someone chosen out of phonebook than a graduate of Harvard.” I’m riffing on that. There’s a sense here where it’s like the wisdom of the crowd, the wisdom of the median person…

31:56 Matthew Feeney: But it’s also though not necessarily wisdom of the crowd that seems to be, it’s wisdom of people who are interested or motivated to be involved. So if I get in a fight with you about… I don’t know, who the Pope was in 1710, we can just whip out our phones and we’ll probably go to Wikipedia, which isn’t anyone we know, it’s just curated by people who do care about who the Pope was in 1710. We can check the logs and see editing, but it’s considered fairly reliable, right? Because it’s entirely built around by a community of people who care about what’s going on without caring about whether Matthew’s right or Paul’s right.

32:33 Paul Matsko: Yeah. Well, and then here’s the recourse to arbitration system. So in those, that small percentage of cases where it’s actually not accurate, where Wikipedia was some malicious editor put something false on there or just got it wrong. I mean, I have to note that as a former history teacher, I would have docked someone’s paper if they put in the footnotes Wikipedia. But…

33:00 Matthew Feeney: Yeah, it’s good for bar fights.

33:01 Paul Matsko: It’s great for bar bets. Or in the case, I’m trying to imagine now that sports betting is nationally legal online or could be, depending on how states rule individually, but imagine the end of a football game, NFL​.com records that a touchdown was caught, but then the referees decide to review it and after a five‐​minute delay, they reversed that call. Everyone’s always got a lot on the line in that moment. But imagine if you’re a smart contract automatically enforced when the six points went up, meaning that you met the bet, you met the over‐​under and then suddenly, oop, it’s reversed but too late, the contract executed. But again, that’s not a super frequent happening. That’s something that happens in a very small percentage of cases, Wikipedia being wrong. This small percentage of cases, that’s the point of this private arbitration system is for dealing with that less than 1% of situations where something messes up.

34:08 Paul Matsko: I thought it was interesting, Kate, when you talked about trust in smart contract. Smart contracts as a way of minimizing trust. I think in my own head, I don’t think of smart contracts as a means of eliminating trust, like in an idealized world where you could make a completely trustless smart contract, I actually don’t know if that’s the goal. It’s almost as if there are some contexts in which making certain things less ambiguous will actually allow you to build more trust on the things that need to remain ambiguous, like making some stuff ironclad, means that other things you can trust people more on. You’ll have to spread the trust around. I was thinking here of… Well, the counter argument would be what I call the prenup effect, the prenup agreement for marriage. All the social science data shows that when you sign a prenup, the moment you sign it you are significantly more likely to get a divorce. So, ironically, making something in that marriage contract more ironclad, less ambiguous, more specific actually decreases the amount of trust in the relationship.

35:24 Paul Matsko: So sometimes making something less ambiguous can actually undermine trust but at the same time, we sign a marriage contract as a means of building trust. So, I’d like to hear a little more about that. Where do you see that balance between smart contract as a means of decreasing the amount of trust required in a contract and smart contracts as a way of actually building trust between the participants by decreasing ambiguity?

35:56 Kate Sills: Sure, yeah. So like I said, I think there are number of different contracts, and they’re all used in different situations. In the case of a marriage, it’s not like you’re making a contract with a stranger, you hopefully have had a very long relationship with this person and you trust them. So I think definitely human psychology comes into play, where you have this person that you wanna spend the rest of your life with in that situation, and then all of a sudden you’re putting in terms that turned into an arm’s length business transaction. I think that itself is probably very disturbing to people, is that kind of transition from what would be a non‐​transactional relationship into a transactional relationship. But I think, again, what I see the use of smart contracts for is not necessarily these contracts that deal with already existing relationships, but rather people who are interacting, who have no other information about the other person, and they’re trying to do business through the Internet.

37:13 Kate Sills: So I think that’s really the primary use case for smart contracts is not to take a real life relationship and turned into a business transaction, but to enable business transactions that would have never have happened without a smart contract. So it’s similar to what economists will say about the market in general. It’s not that we want to have parents buying and selling things with their kids or something like that. We’re not trying to replace human relationships, it’s that the market allows us to produce welfare for people that we wouldn’t otherwise care about, that live on the other side of the world, and we’re actually enriching their lives through business transactions that wouldn’t have happened without a secure marketplace and secure transactions.

38:04 Paul Matsko: And to use Thomas Haskell’s formation, he talks about market capitalism more generally as like a technology of an idea. It expands the circle of who is my neighbor. So it’s not as much about deepening already existing relationships in your village or in your country, it’s about expanding the circle of who you’ll have a relationship with. Now, because of market capitalism, a merchant in Edinburgh in Scotland can have a relationship with a merchant in Bombay, and have some expectation of that relationship actually binding the two together in a way that wasn’t as true in a pre‐​market era. So it’s about widening the pool rather than deepening the pool, in a sense. And that’s how I think I’d apply it here.

38:57 Paul Matsko: Well, we’re running out of time here. I think if I had to sum up one of the interesting themes that’s run throughout would be that, Kate, you’ve been pointing on that there’s nothing new under the sun, that smart contracts are not this last couple of years a product of cryptocurrency in the blockchain, that they predate it, they go back to the early ‘90s, that the idea of independent arbitration courts are older, they go back to the Middle Ages as Tabarrok pointed out, and thus… I think that tempers our expectations here, where we say, “Look, these are old forms that are being enabled to be used in new ways because of new technology.” I think that’s a valuable insight. So Kate, thank you for coming on, thank you for talking about smart contracts with us. I appreciate your time. And listeners, we’ll talk to you next week on Building Tomorrow.

39:57 Paul Matsko: Building tomorrow is produced by Tess Terrible. If you enjoy our show, please rate, review, and subscribe to us on iTunes, or wherever you get your podcast. To learn about Building Tomorrow or to discover other great podcasts, visit us on the web at lib​er​tar​i​an​ism​.org.