クリティカル・ポイント・ポッドキャスト・エピソード3：SheaとAndersが語るブロックチェーン（パート1）(英語版のみ）

クリティカル・ポイントの本エピソードの中で、ミリマンのコンサルタントであるShea ParkesとAnders Larsonは、ブロックチェーン技術についてまだよく知らない方に向けて、ブロックチェーンについて簡単に紹介しています。ブロックチェーンに関する一連のエピソードの第1回目は、ブロックチェーンおよび保険への応用に関する踏み込んだ会話の場面を設定するため、（このテクノロジーが実際に何であるか、また、暗号通貨にどのように関わっているのかを含む）ブロックチェーンの基本について掘り下げます。本シリーズの今後のエピソードでは、保険、アンチパターン、代替手段での利用可能性についてもお伝えします。

Transcript

Disclaimer: This podcast is intended solely for educational purposes and presents information of a general nature. It is not intended to guide or determine any specific individual situation, and persons should consult qualified professionals before taking specific action. The views expressed in this podcast are those of the speakers and not those of Milliman.

Anders Larson: Hello, and welcome to Critical Point brought to you by Milliman. I’m Anders Larson and I’ll be your host today. This episode of Critical Point we’re going to be talking about blockchain and starting a series of podcasts on blockchain and potential applications in insurance. Today, we really just want to start with an introduction to blockchain. So no insurance applications today. Just getting to know blockchain. We think Milliman has a unique perspective on this specifically as it relates to insurance. We work with all stakeholders in the insurance industry. We also have experts from the technology space. So I think we can kind of come at this from a unique perspective. So today, we’re here with Shea Parkes, principal of Milliman. And we’ll just get right into it. I’m coming from this as a perspective of somebody who does not know a whole lot about blockchain, so we’ll just start with the basics. Shea, what is blockchain? Let’s bring it out of the abstract and talk about what people mean when they say blockchain.

Shea Parkes:Well, people mean all sorts of things when they say blockchain and, really, they’re all true. When you get to down to a definition of it, it's an algorithm, it’s an idea, it’s a data structure, it’s a tool to support other technologies and, really, it’s all of those things. You know, if you want to trace it all the way back, there was—it was first an idea and it was an idea published by an anonymous author. You can go look up the original blockchain and bitcoin article and learn a little bit about it there, but, yeah, it’s everything from an idea to a data structure.

Anders Larson: Yeah, and for me and probably for a lot of our audience, it is kind of this idea and you hear “bitcoin” and that tied together and I want to get into all that today of how that actually relates to each other, but let’s start with kind of the basics of a blockchain. What is stored on a blockchain?

Shea Parkes: Okay. So most blockchains right now are used for cryptocurrencies like bitcoin and other ones. And so, that the classic blockchain algorithm, what’s stored on it, is just a series of transactions. So each transaction is, hey, User A sends two bitcoins to User B, and then User E sends five bitcoins to User W and just this long, long history of who’s sending bitcoins to whom. And what it’s often used for is you take that long history, roll it up, there is an inception, there’s a way to get a few bitcoins at the beginning, and along the way—and you can use it to figure out how many bitcoins all of these anonymous users have. And that’s—I said “bitcoins” and I’m going to catch myself a few times here because we’re really talking about blockchain, which is a technology, a tool, a data structure behind cryptocurrencies like bitcoin, but it’s really hard to sort of talk about a blockchain without dragging up some of that baggage that comes with bitcoins and cryptocurrency, so at its base, it’s a log—transaction log of movements of assets.

Anders Larson: And, again, just to make this point, it’s critical that it is a chain of transactions. It's not a group of unique individual transactions. You have to have the whole chain.

Shea Parkes: Yeah. The whole thing. There’s an ordered list of it, yeah. So you can see the exact order that everything happened and that way-- part of the algorithm actually does make sure that, Anders, if you’re representing User A and you only have five bitcoins, you can’t send six or seven bitcoins to someone else. Those transactions would have been rejected from the chain because there is an order to them. So you can figure out account balances at any given time or, more importantly, ownership. If you’re leaving behind cryptocurrency, there’s this abstract asset of something and who’s owning it at any given time. There’s one unique owner at any given time.

Anders Larson: So sort of coming back to cryptocurrency and let’s just say bitcoin, because I think that’s what primarily most people know of when they think of cryptocurrencies, what makes this such a good fit and why is it so interrelated to cryptocurrencies?

Shea Parkes: Right. I think part of it is that it’s really good at tracking a ledger of the ownership of something. And when that ownership is a made-up currency, it’s a really, really nice fit there because you don’t have to have anyone else involved to sort of prove that ownership or anything. That blockchain, that list of transfers from those users, that’s all that’s needed. There’s nothing else. Like I can purely prove I have so many bitcoins just based by that transaction history out there. And the other was sort of an ideological goal of bitcoin and the other cryptocurrencies, that they didn’t want to report to a central bank or a government, or they wanted this currency that was outside of control of almost all current control figure. And the blockchain algorithm let them get there. But it’s really hard to get a whole ragtag group of not misfits, but ragtag group of people you generally don’t trust. People participating in the bitcoin blockchain and they participate by donating their computing resources and such, they don’t actually trust anyone else on the network. They don't trust anyone else working with bitcoin, and so how do you make an algorithm that people can trust in the algorithm while they’re not trusting any of their peers. And that’s some of the real sort of breakthrough in the blockchain algorithm.

Anders Larson: So there’s no bitcoin blockchain honor code.

Shea Parkes: No, oh, no, no, there’s no bitcoin honor code. Like, if you mess up or if someone gets to your secret keys and takes all your money, there is no power to appeal to. That money is gone.

Anders Larson: All right. Let’s talk about one of the terms I’ve heard bandied about related to that topic is immutability.

Shea Parkes: Yeah, so, immutability, it’s a term and a concept a lot from computer science. It means kind of like “read only” or “write once” as in like, once that transaction has been recorded, you can’t change it. The transaction is frozen in history, all that-- so, the concept of freezing, unchangeable, it’s sort of summed up in the term “immutable.” The transaction log of a blockchain is immutable, which is actually kind of hard to get to. It’s not hard to get to when there’s a central authority if you trust that authority. That’s sort of the whole issue in blockchain and cryptocurrency was “Which central authorities do you trust?”

Anders Larson: So as we were talking about this, preparing for this show, a lot of this has to do with the chaining nature, but then there were kind of what you described as math problems or kind of...

Shea Parkes: Puzzles.

Anders Larson: ...puzzles that have to be solved and they have to be solved sequentially in order for the chain to be essentially hacked and sort of for the immutability not to work. Could you maybe kind of start us from scratch about...

Shea Parkes: Sure.

Shea Parkes: Yeah, so blockchain is actually a very blunt name. Blockchain is a data structure that is a chain of blocks. Each block is pretty dull actually. It’s just a series of transactions of, you know, I said, two coins from User A to User B, five coins here and there, yonder. That’s the transactions that go into a block. Now, you can participate in a blockchain in a few different ways. You can actually be running the computer resources yourself. You can have all this transaction log on either a physical computer, virtual computer that you’re in control of, or you can have someone else running it and you’re actually just sort of-- just talking about how you interact with it that way. But let’s pretend you’re running it yourself. Well, then you’re participating in this web of a blockchain and you’re saying, “Hey, we want to agree on what the next block will be.” So we have people participating sort of submit what transactions they’ve heard about. Maybe Anders is both participating and that he has some bitcoins and he’s running the bitcoin blockchain software on his laptop or something. He’s not, but he could. Well, not his personal laptop. All this should apply to your own company, DUAs.

Anders Larson: I think we’re safe there.

Shea Parkes: I think we’re safe there. But he could be running the block, participating in the bitcoin blockchain network on his personal laptop. He could be putting in his requests for a transaction to happen. Once those requests sort of get bundled up for a certain time window-ish, then the whole world starts racing to solve a math puzzle. And we can talk more about that math puzzle because it is a pretty important part to understand those math puzzles and how they might apply in other cases, but for now think of it as a math puzzle that you really just have to brute force. So pick—it’s going to have to be a number between one and a billion, and when you pass it through this formula, the shuffley formula, the answer needs to be less than a thousand. So it’s a brute force puzzle. It takes maybe a minute to try a number to see if it works. And maybe only one number in a million or one number in a billion actually work. So everyone around the world’s trying to compute this puzzle. They’re all guessing numbers—they’re called mining. They’re trying to figure out what numbers solve this puzzle and as soon as they solve this puzzle, they announce to everyone else, “I found it, I found it.” And sometimes—well, first of all, they found that puzzle. Everyone else can verify that their solution worked because it’s not so hard to verify it that one time. They run the number through. It’s hard instead to find that one number in a million numbers, right? So everyone can verify it so that those numbers then become bundled into a block and then it becomes the next block in the chain. And the person who contributed to that block and solved the puzzle, in the classic bitcoin implementation of blockchain, they might get a bitcoin sometimes for adding the next blocks. That’s one place those bitcoins can sort of have an inception point. But what I think you asked rightfully is, “What does all this have to do with immutability?”

Anders Larson: Right.

Shea Parkes: Well, immutability means I can’t go back and change history, is where I was going here. And so, those histories are in blocks that happened an hour ago or a day ago or a week ago. So if I go back and change—if I want to change a transaction that happened, like, a week ago—

Anders Larson: If you want to steal it and say, “I paid you,” instead of “I paid...”

Shea Parkes: Or more exactly, like, “Hey, I transferred Anders... I went...” Say I’m “A” and Anders is “B.” I transferred 10 bitcoins to Anders and he gave me a car. I got the car, now I drove away, now I want to go back and change the history to where I never actually gave the bitcoins to Anders, so then I end up with a car and the bitcoins. That would be a malicious actor in this scenario.

Anders Larson: It would be.

Shea Parkes: While I can go back and change that transaction of that block, but that changes the puzzle. So I need to resolve that puzzle, but then the chain part of the block says that each puzzle for each block also depends on the answer to the puzzle in the last block. So I changed the puzzle for the block from a week ago, when I paid Anders for that car, I have to resolve it, but just on my own. The rest of the world is not trying to help me here. Just me trying to solve this new puzzle, then I have to solve the puzzle for the next block, then I have to solve the puzzle for the next block, and everything up through current, because one neat thing is the world—like, the algorithm consensus is they believe the longest chain. And so, I’ve changed something in the past. I’ve got to somehow get my chain longer than the global world chain. And I’m really going to lose because the rest of the world is still contributing and chasing and racing these puzzles and I’m not going to have enough compute resources to solve puzzles faster than the world. So my chain will never catch up with the world. It will blow me away. And so, that's part of the immutability of these sort of puzzles that are out there, that I can’t solve puzzles faster than the world.

Anders Larson: Right. And that’s why as we talked about at the beginning, it sort of started to crystalize a little bit for me, was the chaining nature. And that’s why it’s so critical that this log is not a bunch of individual transactions that could just be sorted in whatever order. It’s the chain nature is critical to the immutability because it has to be in that order, and then the puzzling having to be solved in that order is what provides the immutability and really the safety from hacking.

Shea Parkes: Yeah. And there is a risk that if some malicious entity physically bought two-thirds of the computing power in the world, then they could race and re-write history. But that is thus far not played out. Although there are some funny things that happen in the weird economics of custom-built hardware and things that have come close. But for the most part, no one’s able to beat the world. And so, that’s the protection. That’s how the classic cryptocurrency blockchains get their immutability and how people trust in the algorithm instead of trusting in their bank, even though banks offer things—niceties as well. We’re not in any way, shape, or form saying that you should take your money—yeah. This is not a suggestion to take your money out of a bank and go buy a bunch of bitcoins. Talk to your financial advisers.

Anders Larson: This is just an explanation of why the system has been working.

Shea Parkes: And what its goals were. Because its goal was they wanted to make it to where there was no central authority to chase these bitcoins. Well, cryptocurrencies: no central authority to control these cryptocurrencies.

Anders Larson: So, I think, you know, we talked a lot about cryptocurrencies today. We talked a lot about bitcoin in particular. But I think it’s helpful to understand how blockchains work in the real world as we then transition into some of our future episodes, potentially to applications in insurance, which maybe are not immediately apparent from what we’ve discussed today, but I think the foundations hopefully we’ve laid so that if you’ve listened to this episode, you can come into the next episode and we can start talking about how this transitions to potentially the insurance industry.

Shea Parkes: Yeah. And maybe one thing we left off here was, you know, again, don’t take your money out of a bank, but if you did want to buy something like a bitcoin, what’s the mechanism? And you can run the software yourself and maybe occasionally earn a bitcoin for solving puzzles. In all honesty, you’re not going to win very many because there’s a lot of people out there on most of the popular cryptocurrencies doing this. Usually, what would happen is you would have to give a real-world—is a bitcoin cryptocurrency a real-world asset? I don't know. Is a dollar a real-world asset? I don’t know. Those are kind of deep questions. Let’s just say you have to convince someone that has bitcoins that you give them something like a car or maybe transfer some money via PayPal or something.

Anders Larson: Right.

Shea Parkes: And convince them to put a transaction into the blockchain of bitcoin that moves some of their bitcoins to you, for some definition of you, which that’s a whole other topic of another podcast probably is how do we know who each other are on a blockchain? Because that becomes really, really important in the purpose of any sort of application of insurance or otherwise.

Anders Larson: Okay, so, to kind of start wrapping up, let’s say people want to learn more—again, not necessarily advising getting into the bitcoin world, but if you just want to learn more about blockchain, more about bitcoin, what are some resources for people out there?

Shea Parkes: All right, well, you can definitely, you know, internet searches are your friend, lots of great topics out there, you can find the original article. It’s a highly technical article. My favorite resource that sort or really helped me understand the data structure and the algorithm was a book known as “Blockchain Basics” by Daniel Drescher. It was a non-technical introduction in 25 steps. I felt it did a really good job of laying out the goals of blockchain, how those goals align with cryptocurrency, but still addressing the blockchain independent of cryptocurrencies and it also does a good job of explaining what was this breakthrough? What was blockchain specifically?

Anders Larson: Yeah, and I think, you know, this podcast today, our first episode—really been about blockchain specifically. We have talked a lot about cryptocurrencies and bitcoin in particular. We know that Milliman, our focus a lot of times is primarily on insurance and we want to get this back to the insurance focus, but we thought it was really critical to have this primer to kind of get everybody up to speed on what blockchain is so that we can kind of dive deeper into potential applications in the insurance industry.

Shea Parkes: Yeah, or existing applications. There are actually already some out there.

Anders Larson: Right. So thank you for joining us. Thank you, Shea Parkes, for talking with us today. You’ve been listening to Critical Point presented by Milliman. To listen to other episodes of our podcast visit us at milliman.com or you can find us on iTunes, Google Play, Spotify, and Stitcher. We’ll see you next time.