Lodestar：資料留鏈下，誰來證明沒造假？
The Root of Trust: A Case for Optimism in Ethereum's Future
- Matthew Keil ChainSafe／Lodestar
影片
影片與逐字稿來源：ETHTaipei
場次摘要
Lodestar 工程經理 Matthew Keil 主張，應從使用者的信任問題設計產品，再決定以太坊要負責哪一段。應用邏輯和資料可以留在伺服器及資料庫，只把需要公開驗證的狀態交給鏈。他以投票和棉花供應鏈為例：可驗證不等於可以公開所有資訊，隱私與防脅迫也必須納入設計。
值得記下的 5 件事
- 區塊鏈數位化的是讓金錢運作的信任；金錢只是這份數位化信任的第一個應用。
- 產品架構應從問題出發，再決定以太坊負責哪一部分；應用邏輯與資料不必全部上鏈。
- 混合架構可把邏輯留在伺服器、資料留在資料庫，只把存取紀錄與可驗證狀態連回以太坊。
- 投票者若能證明自己投了什麼，就會打開買票與脅迫的空間，投票系統不能只追求可驗證。
- 棉花供應鏈概念驗證未獲採用，因為參與者不願揭露供應商、交易價格與採購頻率等競爭資訊。
逐字稿
區塊鏈數位化的是信任
今天我準備了一個有點辛辣的題目，更重要的是，我想對這個領域的建設者發出行動呼籲。我是 Matthew Keil，Lodestar 團隊的工程經理。Lodestar 是為鏈提供安全的共識客戶端之一，我們做的是權益證明。好，下一張。信任被打破了。發明區塊鏈技術的那些人，對印鈔票的人失去了信心。So today I have a little bit of a spicy topic, and I have, more importantly, a call to action for the builders in the space. So my name is Matthew Keil, and I'm the engineering manager of the Lodestar team. Lodestar is one of the consensus clients that provides security for the chain. We do the proof-of-stake. Yeah, I'll just—okay, next slide. Trust was broken. The people who invented blockchain technology lost faith in the people that print our money.
「Chancellor on brink of second bailout for banks.」有趣的是，我們失去信心的不只是那些人，我們連對錢本身都失去了信心。好，往右。錢這東西很妙：錢是一種幻覺。只有大家同意它是錢，它才會變成真的。你跟我都把那張紙看成錢，它就變成黃金。就是這種共同信念，把紙或塑膠，變成對社會更關鍵的東西。Chancellor on brink of second bailout for banks. The funny thing, though, is it wasn't just the people that we lost faith in; we lost faith in the money itself. Got to go with the right. All right. And the funny thing about money: money is an illusion. It's something that only becomes real when we all agree. If you and I both see that piece of paper as money, it becomes gold. And it's that shared belief that transforms it from paper or simple plastic into something more critical to society.
區塊鏈是數位化的信任。它把同一種信念、同一種信任數位化，這才讓這項技術特別。我們數位化的最根本東西不是錢，是讓錢能運作的那份信任。錢是這份數位化信任的第一個應用，但底層那個原語重要得多。往下之前，我想先問大家一個問題。請幾個人舉手。什麼是信任？Blockchain is digitized trust. It's the digitization of that same belief, that same trust, that makes this technology special. The most fundamental thing that we digitized was not the money; it was the trust that makes it work. And while money was the first use case of this digitization of trust, the underpinning and the primitive is much, much, much more important. So now, before we go any further, I want to ask you guys a question. I'd like a couple of people to raise their hands. Who wants to raise their hand? What is trust?
誰要回答？什麼是信任？好，你先來。我覺得信任是你相信對方不會侵害你。好，這是一般的感覺。對，就是信任。你知道對方有能力侵害你。這也是一部分。還有誰？但你相信他們不會那樣做，那就是信任。這很有意思。注意，你在定義這個東西時，用了「信任」這個詞本身。這是我想請大家帶走的一點。Who wants to give an answer? What is trust? Yes. What is trust? Go for it first. Yes. What is trust? I think trust is something that you believe the other side won't violate you. Okay. That's the general sense. Well, it's trust, right? So you know they have the capability to violate you. Okay, also part of it. All right, who else? But you trust them that you won't be violated. That is trust. And this is a very interesting thing. Notice that you use the word trust in the definition of the thing itself. And this is the part I want you to take from this.
從問題出發，不必全上鏈
我問「什麼是信任」時，我保證你骨頭裡就感覺到了，你立刻知道信任是什麼。但一要講成話，就變得比較糊、比較複雜。我覺得這正是現在這個生態被困住的根本之一：我們太專注在自己要蓋什麼，反而看不見自己站在上面的那項技術有多重要。好，這是我對信任的答案：信任是我們衡量誠意的能力。When I asked that question, what is trust? I guarantee you felt it in your bones. You knew immediately what trust was. However, when you go to put it into words, it's a little bit more muddy. It's a little bit more complicated. And I think this is one of the fundamental things that is plaguing our ecosystem today: we are so hyper-focused on what we're trying to build, we're losing sight of the importance of the technology that we're building on top of. Okay. Now, this is my answer to what trust is: it's our ability to gauge sincerity.
我們要不要跟另一個人、另一個機構往來？我們要怎麼宣布「我是好夥伴」？我想把設計空間打開的，就是這件事。你們接下來要找題目做時，請把這種感覺帶走。區塊鏈是架構工具箱裡的一種工具。好的架構師從問題出發，不是從自己想用的技術出發。舉例，我不會先說「我要用 Kafka 來做」。Do we want to engage or participate with another individual, another institution? Do we want to—how do we announce that I'm a good partner? And this is where I'd like to open up the design space. It's this sentiment that I want you to carry with you when you're looking to explore what is next to build. Blockchain is a tool in the architectural toolbox, and a good architect starts with the problem, not the technology that they want to solution for. As an example, I wouldn't say, "I want to build with Kafka."
然後整個設計都圍著 Kafka 轉。我會先問：push 是不是比 pull 更合適？我要做發布訂閱，還是只要輪詢就能拿到需要的資訊？我希望你們用同一種方式看區塊鏈。我挑戰你們：圍著問題做架構，去用以太坊真正擅長的那一塊。重點就在這裡。不是每件事都要上鏈。我知道這在以太坊社群是辛辣題目。但不是所有應用邏輯都要上鏈，也不是所有資料都要上鏈。And then start designing around the use of Kafka. I would first ask the question, is push a better technology to use than pull? And do I want to do pub-sub, or do I want to just do polling in order to get the information that I need? And this is the same way I want you to view blockchain technology. I challenge you to architect around the problem. And leverage what Ethereum does effectively. And this is where leveraging Ethereum is important. Not everything needs to be on-chain. I know that's a spicy topic in the Ethereum community. But not all application logic needs to be on-chain. Not all data needs to be on-chain.
反直覺的是，我也不認為所有資料都該完全透明。真正重要的架構問題是：以太坊需要在哪裡當信任根？想像一個其他方面很普通的應用，邏輯在伺服器上，資料在資料庫裡。但這個資料庫很特別。它會把資料庫存取回報到區塊鏈，讓你知道是誰、在什麼時候讀了這份資料。你可以做成事件，而支撐這個資料庫的資料結構本身是 Merkle 化的結構。And counterintuitively, I don't think all data should be fully transparent. The important architectural question is, where is Ethereum needed as the root of trust? Now, imagine an otherwise conventional application, one where application logic lives on a server and data lives in a database. However, the database is special. This database has blockchain reporting for access to the database so that you know who and when people are reading from that database. You can event it, and the data structure itself that underpins the database is a Merkleized structure.
投票與供應鏈的混合架構
你可以做出可證明的狀態，可以看到所有交易，這就給了你我說的信任根。這聽起來很熟，是因為我們從 rollup 的想法借來的：不是所有應用邏輯都活在鏈上，它活在別的地方，安全保證再捲回以太坊，讓信任根灌注到其餘流程。一個很好的例子是投票系統。ZK 投票有先天缺陷，這也是我們今天不用它的原因。問題是，你沒辦法解掉「我怎麼避免賣掉自己的票」。You can create provable state, you can see all transactions, and this gives you the ability to do the root of trust that I'm talking about. And if this sounds familiar, it's because we're borrowing it from the idea of rollups. Where not all application logic lives on-chain. It lives somewhere else, and the security guarantees are rolled back up to Ethereum, where the root of trust can be imbued on the rest of the process. And a perfect example of this is building a voting system. ZK voting has inherent flaws. It's why we don't use it today. The problem with it is that you can't solution around how I do not sell my vote.
投票的關鍵之一是：你不該能證明自己投了什麼。但在 ZK 投票裡，我會拿到 witness。這就打開共謀、脅迫、買票的門。這是我們今天不做 ZK 投票的原因。用這種混合架構，你可以繞過這個問題，讓區塊鏈做它最擅長的事。不必等月球數學追上我們夢想的未來，我們可以為今天的現實來做。這不是，好，再一張。混合架構不只這一個用途，供應鏈是另一個很好的例子。One of the critical pieces of voting is that you should not be able to prove what you did, and I will have the witness in ZK voting. So it opens the door to collusion, coercion, and vote buying. This is why we don't ZK vote today. However, using this mixed architectural pattern, it gives you the ability to solve around the problem and use the blockchain for what it's best for. So instead of waiting for moon math to catch up to the future we dream of, we can build for the reality of today. Now, that is not—oh, one more. Okay, that is not the only use case of one of these mixed architectures, and supply chain is another great example.
我幫棉花產業做過一個概念驗證。他們需要證明棉花來自這個國家、或世界的某一區，而不是另一區。最後沒被採用的卡點是：沒有人想洩漏自己的競爭優勢。在組織內部看供應鏈，貨物流進流出很容易看懂。區塊鏈真正有用的地方，是組織與組織的邊界，東西從一家公司交到下一家，而競爭優勢就在那裡。我不一定想說我跟誰買、什麼價格、什麼頻率在買。So I built a proof of concept for the cotton industry. They needed to prove that cotton came from one part of the country or one part of the world and not another. And the rub that caused it to not be used is that nobody wants to spill their competitive advantage. When you look at supply chain, it's very easy within an organization to understand where the flows of things go to and from. However, where blockchain really becomes useful is at the boundary between organizations, when things move from one company to the next, and that's where the competitive advantage is. I don't necessarily want to say who I'm buying from, or what price I'm buying at, or at what frequency I'm purchasing.
所以沒有人想採用那套系統。我做出來的東西沒有踩在現實上。我覺得最重要的就是：為今天的現實來做。我們可以等數學追上，可以等系統更有效率、等 ZKVM 更即時。但一個沒有人會用的理論系統，是沒用的。隱私跟以太坊並不互斥。受信任的元件不是區塊鏈的敵人。把區塊鏈用在 DeFi 以外的地方，也會解掉我們最大的問題之一。This is why nobody wanted to adopt the system. I built something that wasn't based in reality. And this is the part that I think is most important: building for the reality of today. We can wait for the math to catch up. We can wait for systems to be more efficient, for ZKVMs to be more timely. But a theoretical system that no one will use is useless. Privacy is not incompatible with Ethereum. Trusted components are not the enemy of blockchains. Using blockchain for more than DeFi will also solve one of our biggest problems.
對外信任問題與建設呼籲
我知道這不是今天現場建設者的焦點，但這正是我今天要講非技術題的原因：我要呼籲建設者去做能解區塊鏈最大問題之一的技術零件。我們有信任問題。外面大多數人不信任區塊鏈。我覺得這正在拖累我們這麼喜歡的這項技術被採用。鏈上有太多價值抽取、迷因幣，以及世界上最糟的那一面。這就是讓我媽不想用它的原因。所以我今天的行動呼籲在這裡：信任可以被重建。And I know that is not the focus of the builders here today, but this is exactly why I'm giving this non-technical talk today: I want to call to action for the builders to build the pieces of technology that will solve one of the biggest problems that blockchain has. We have a trust problem. Most of the people in the outside world do not trust blockchain. I think that is hampering the adoption of the technology that we love so much. There are a plethora of examples of people doing value extraction and meme coins, and just generally the worst parts of the world just on-chain. And this is the part that makes my mom not want to use it. So this is where my call to action comes in today. Trust can be restored.
我們可以用這項技術做很厲害的事，不管是投票、供應鏈、憑證管理、保險商品、房地產投資。政府課責我覺得是潛力最大的一塊。今天真正該帶走的是：被看著的人不想被看。政府會很猶豫去為自己做監督。但我們可以做。用這些混合架構，或建立在區塊鏈技術上，去創造我們當初為了解金錢問題而創造的那份信任，可以解掉很多其他問題。We can use this technology for amazing things, whether it be voting, supply chain, credential management, insurance products, real estate investment. Government accountability, though, is one of the things where I think it has the most potential. And this is the part that is really, really critical for us to take away here from today. The people who are being watched do not want to be watched. Governments will be hesitant in order to create their own oversight. However, we can do that. Using these mixed architectural patterns or building on blockchain technology in order to create this trust that we created in order to solve the money problem can solve so many other issues.
這就是我今天對你們的挑戰。想開一點。把區塊鏈當成能創造更平等世界的架構解法。信任被打破了。我們做出了能修它的工具，我挑戰你們去找一個值得修的問題。有問題嗎？我有一個問題。好。先謝謝你的演講。謝謝。有問題請舉手，儘管問。And this is where I challenge you today. Think outside the box. Use blockchain as an architectural solution that can create an egalitarian world. Trust was broken. We built tools that will fix it, and I challenge you to find a problem worth fixing. Questions? I have a question. Yes. All right. Well, first of all, thank you for the talk. Thank you. Is there any questions you'd like to ask? Raise your hand. Feel free.
希望你們有被啟發。今天不是很技術。有被啟發嗎？我有一個問題。好。以你現在的看法，值得修的問題是什麼？有好幾個。我覺得最重要的之一，其實已經開始在 Twitter 冒出來：客戶端開發、以及鏈本身開發，前面已經看得到資金問題。這裡也有信任問題：我們信任社群會資助公共財，社群也信任我們會去做，只因為我們「應該」做。Hopefully you're inspired. Not very technical today. Are you inspired? I have a question. Yes. Yeah. So, what do you find the problem worth fixing, in your opinion, right now? There are several, and I think one of the most important is actually starting to bubble up on Twitter. There is funding issues that are on the horizon when it comes to client development and development of the chain itself. And this is where there's also a trust issue, is that we trust the community is going to fund the common goods, and the community trusts that we are going to do it because we just should.
問答：公共財資金與沒被計算的票
我們很多人，我自己也在一個很棒、以做以太坊為榮的團隊。事實上我看到另一位先生，他也以做以太坊為榮。我們願意免費做。但不是每個人都這樣。我覺得最重要的信任問題之一就在這裡：做出鏈上治理管理系統，讓資金能以更平等、更看貢獻的方式，流到以太坊的開發上。這是我開始在做的事。是。你的演講讓我很受啟發，很有意思，也很鼓舞人。And while many of us, and I work on an amazing team of people that is proud to be working on Ethereum. As a matter of fact, I see another gentleman who is also very proud to be working on Ethereum. And we would do it for free. But that's not true of everybody. And this is where I think one of the most important trust issues comes in: building an on-chain governance management system that allows for funding to be able to flow in an egalitarian way, a meritocratic way, towards the development of Ethereum. It's something that I am starting to work on. Yes. I was quite inspired by your talk. Very intriguing. And very inspiring.
但開頭我不太懂。你說信任被打破了，又挑戰我們去找問題來做。現在這個時代，信任被打破的好例子到底是什麼？我給你一個最貼切的例子，也是我把投票放上來的原因。我是美國人，如果這還不夠明顯的話。我住在海外，上一場選舉我把票傳真回去。傳了三次。我拿到傳真成功的確認，所以我知道它到了那台機器，但我的票從來沒被計算。But in the beginning, I don't really understand what is the—you said trust was broken, and you challenged us to find problems to work with. But what exactly is, like, a good example of a broken trust problem in our days? I'll give you the perfect example and the reason why I put voting up here. So I'm an American, if it wasn't apparently obvious. I'm an American, and I live overseas, and I faxed my vote in for the last election. And I faxed it three times. I got verification that the fax went through, so I know it got to that machine, and my vote was never counted.
你怎麼知道它沒被計算？可以上網查。有選民紀錄，上面基本上寫我的票沒被計算。我知道我寄了，我有寄出的證明。而我住的那個選區，是我沒投的那個人以非常些微的差距贏。對我來說，這是腐蝕民主的力量。如果我的票不會被計算，我為什麼還要投票？謝謝。這讓我想起，我確實有類似經驗。謝謝。好，還有其他問題嗎？最後一個？How do you know it is not counted? You can look online. Like, there's a voter record, and it basically said that my vote was never counted. I know I sent it. I have verification that I sent my vote, and it was a very slim win in that district that I live in, for the person I did not vote for. And this, to me, is a caustic force on democracy. Why should I vote if my vote doesn't get counted? Yeah, thank you. So it reminds me of—yeah, I do have some similar experience. Yeah, thank you. All right, any other questions? One last one?
誰想去做能改變世界的東西？所有人。我知道我想。我們去改變世界。我很確定這裡每個人都會。我們做，一起做。好，謝謝。Who wants to go build something that changes the world? Everyone. I know I do. Let's go change the world. Pretty sure everyone here is going to. Let's do it. Let's do it together. Let's do it. All right, thank you.
