比特币如何重新定义了一个经典的计算机科学问题

那是2007年的事了。

So this was in 2007.

我当时参加了一个研讨会，主要想看看拜占庭容错到底有没有实际应用价值。

I was at a workshop, and the goal of the workshop is kind of see whether Byzantine fault tolerance is practical or not.

但当时有两个主要的质疑。

But there was actually two big complaints.

一个是说根本没人需要这个技术，另一个是说性能太差了。

One is that maybe nobody needs it, and the other was that the performance was horrible.

所以问题根本不是共识协议能不能用。

So the question wasn't whether consensus protocols are practical.

问题是，你真有必要应对那种极其不可预测的故障，而不仅仅是系统崩溃吗？

The question was, do you really need to be robust to potentially very unpredictable failures as opposed to just crashing?

中本聪其实看明白了这一点。

Satoshi Nakamoto, he kind of realized that.

他说，比特币最核心的技术，就是解决拜占庭共识问题。

He said the the core technical aspect of Bitcoin is solving Byzantine agreements.

我觉得，事实上，我们所知的所有主流区块链都在运行某种版本的拜占庭容错机制。

I would say, de facto, all the major chains that we know are running some version of Byzantine thought tolerance.

早期的权益证明协议效率并不高。

The early proof of stake protocols, they were not very efficient.

它们的区块生成间隔长达十分钟。

They had blocks every ten minutes.

如果你真想服务数十亿用户，或者支撑大型经济系统，你就得准备两种模式：和平模式和战时模式。

And so, really, if you're thinking about serving billions of people or systems that really manage large economies, you wanna have kind of a wartime mode and a peace time mode.

在和平模式下，系统不会出现故障。

So in peace time, there's no failures.

但关键是，你必须能随时切换到战时模式。

And the thing is that you do wanna be able to switch to wartime.

一旦遭到攻击，你也有能力应对大规模的系统破坏企图。

So if you are under attack, then you do have a way to kind of overcome a massive attempt to corrupt your system.

人们常常讲比特币的故事，好像它突然从天而降一样。

People often tell the story of Bitcoin as if it appeared out of nowhere.

但比特币背后的理念其实可以追溯到几十年前，根植于计算机科学、密码学和分布式系统的基础研究。

But the ideas behind Bitcoin stretch back decades, drawing on foundational work in computer science, cryptography, and distributed systems.

在本期节目中，蒂姆·罗格加登和伊泰·亚伯拉罕深入探讨了区块链共识的科学源头，解释了比特币为何在拜占庭容错领域实现了突破，并讨论了数十年的学术研究如何持续影响现代区块链协议的设计。

In this episode, Tim Roughgarden and Ittai Abraham explore the scientific roots of blockchain consensus, explain why Bitcoin represented a breakthrough in Byzantine fault tolerance, and discuss how decades of academic research continued to influence the design of modern blockchain protocols.

无论你是刚接触加密货币，还是已经关注这个领域多年，这场对话都能带你更深入地理解支撑现代区块链的科学理念。

Whether you're new to crypto or have been following the space for years, this conversation offers a deeper look at the scientific ideas that underpin modern blockchains.

如果你喜欢这期节目，别忘了订阅 a16z Crypto 播客，获取更多类似的内容。

If you enjoy this episode, be sure to subscribe to the a 16 crypto show for more conversations like this.

大家好。

Hi, everyone.

我是蒂姆·罗格加登，a16z Crypto 研究主管，同时也是哥伦比亚大学计算机科学教授。

I'm Tim Roughgarden, head of research at a16z Crypto and professor of computer science at Columbia University.

今天，我们正式启动一个新系列——‘第一性原理：区块链技术的科学根基’，探讨区块链领域理论与实践交汇处最令人兴奋的研究方向，追溯那些让区块链成为可能的思想源头。

And today, we're kicking off a new series called First Principles, the Scientific Roots of Blockchain Technology that explores one of the most exciting areas of research at the intersection of theory and practice blockchains and where the ideas that make them possible come from.

其核心是几十年来计算机科学、经济学和数学领域的研究成果：分布式系统如何达成共识，信任如何在没有中心权威的情况下建立，以及如何在陌生人的网络中验证计算结果。

At their core are decades of work across computer science, economics, and mathematics ideas about how distributed systems reach agreement how trust can emerge without central authority and how computation can be verified across networks of strangers.

我们会从这些想法的源头开始，一直追溯到如今实际运行的系统。

We'll trace these ideas from their origins to the systems running in production today.

我们还会邀请那些取得关键突破的科学家和学者，和他们深入交流。

And we'll talk with the scientists and scholars whose breakthroughs made it all possible.

首先，我们会聚焦于其中一条最核心的线索：分布式共识——在出现故障或恶意行为的情况下，多台机器如何就共同状态达成一致。

To start, we'll focus on one of the deepest threads: distributed consensus: how many machines can agree on a shared state, even in the presence of failures and adversarial behavior.

像拜占庭协议和状态机复制这样的概念，几十年前就已提出，如今仍是区块链技术的核心。

Concepts like Byzantine agreement and state machine replication, developed decades ago, sit at the heart of modern blockchain technology.

首先，我邀请到了a16z加密货币研究合伙人Ittai Abraham。

To begin, I'm joined by a16z crypto research partner Ittai Abraham.

Ittai是拜占庭协议和共识机制领域全球顶尖的研究者之一。

Ittai is one of the world's leading researchers in Byzantine agreement and consensus protocols.

他还是VMware区块链项目的创始成员之一。

He is a founding member of VMware's Blockchain Project.

他同时也是技术博客《Decentralized Thoughts》的创始人，这是该领域最受尊敬、持续时间最长的博客之一。

He is also the founder of Decentralized Thoughts, one of the field's most respected and long running technical blogs.

我们一起梳理两位先驱者的工作：芭芭拉·利斯科夫和莱斯利·兰波特，接下来的几期节目中，你们会直接听到他们本人的声音。

Together, we unpack the work of two pioneers: Barbara Liskoff and Leslie Lamport, whom you'll hear directly from in the episodes ahead.

好的。

Alright.

所以，Ittai，这真的太棒了。

So Ittai, very, very cool.

我们能采访到莱斯利·兰波特和芭芭拉·利斯科夫两位。

We get to interview, both Leslie Lamport and Barbara Liskov.

这真是莫大的荣幸。

It's a great honor.

能做成这件事，实在太令人兴奋了。

Really, really cool we get to do this.

也许为了让听众更清楚，我们不妨聊聊，他们当年的开创性研究是怎么跟区块链技术联系起来的。

You know, maybe for the audience benefit, we should talk a little bit about how the pioneering work that they did connects to blockchain technology.

你知道，第一个出现的区块链——比特币，2008到2009年左右，它本质上就是一个共识协议。

Know, so the first blockchain that came out, Bitcoin, 2008, 2009, you know, part of what it is is a consensus protocol.

那问题来了，这到底意味着什么？

And so the question is like, what does that mean?

Lamport 和 Liskov 的很多工作，都比比特币早得多。

A lot of the work that, Lamport and Liskop both did, it's way before Bitcoin.

对吧？

Right?

他们的研究早在八九十年代，甚至更早就开始了。

It's from the nineties, from the eighties, even earlier.

那这两者之间有什么关联呢？

What's the connection?

比特币确实是个共识协议，但在此之前早就有经典的共识协议了。

Bitcoin is a consensus protocol, but there's also this classic working consensus protocols.

比特币是在重新发明轮子吗？我们该怎么理解这一点？

Was Bitcoin reinventing the wheel, or how should we think about that?

比特币不只是在分布式计算领域带来了一次全新的颠覆。

Bitcoin was not just a whole new disruption in in distributed computing.

它在经济和密码学方面也有创新。

It also has innovation in economics and cryptography.

但在这里，我们主要关注分布式计算这一部分。

But here, we're gonna focus on the distributed computing part.

人们花了好几年才意识到，这实际上是在解决拜占庭将军问题。

It actually took quite a few years for people to realize that this is kind of solving a Byzantine agreement problem.

拜占庭将军问题长期以来一直是分布式计算领域一个核心的学术难题，研究了整整四十年。

So Byzantine agreement problem was kind of this very core academic problem that has been studied for forty years.

从这个角度看，比特币彻底革新了我们解决拜占庭将军问题的方式。

In that sense, Bitcoin is kind of this huge revolution in how to solve Byzantine agreement.

事实上，如果你看看中本聪早期的邮件，就能发现他早就意识到了这一点。

In fact, if you look at early emails from Satoshi Nakamoto, he kind of realized that.

他曾说过，比特币的核心技术就在于解决拜占庭将军问题。

So he said, you know, the core technical aspect of Bitcoin is solving Byzantine agreement.

所以中本聪显然知道，这其实是一个在分布式计算领域广为人知的问题。

So Nakamoto apparently knew that this was a a well known problem, distributed computing.

那么，比特币协议和我们接下来要听到的那些协议之间有什么关联呢？

So what's the connection then between the Bitcoin protocol and, you know, the protocols we're gonna hear about?

没错。

Yeah.

从某种意义上说，巴芭拉·利斯科夫和莱斯利·兰波特要介绍的这些协议，本质上都是拜占庭共识协议，或者更广义的共识协议。

So so in a sense, both the the protocols that we're gonna hear about from Barbara Liskov and and Leslie Lampart are Byzantine agreement protocols or agreement protocols in general.

但比特币是在一个完全不同的环境下解决这个问题的——无许可、架构更简洁，主要面向加密货币和密码经济学。

So Bitcoin solved this in a much different setting, permissionless with a much smaller setup, much more kind of geared towards cryptocurrency and cryptoeconomics.

但如果从基础原理来看，它们解决的其实是同一个共识问题。

But if you kind of think about this from the foundational perspective, they're both solving the same agreement problem.

共识问题就是指，多个参与方各自有不同的输入，即使其中一部分参与者行为恶意或不诚实，他们也必须达成一致。

So agreement is kind of this problem where you have multiple different parties, and they might have different types of inputs, and they need to reach agreement even though a fraction of the participants are are behaving maliciously or in a corrupt manner.

那么在比特币协议中，哪些参与者需要达成一致？他们又是在同意什么？

So, like, in the Bitcoin protocol, who are the parties that have to agree, and what is it that they're agreeing on?

对。

Right.

你有一群矿工。

So you have a miners.

对吧？

Right?

矿工们通过解工作量证明来生成新区块，推动协议运行。

And the miners are trying to kind of push the the protocol forward by generating new blocks by solving proof of work.

但并不是所有矿工都真心想遵守规则。

And it it might be that not all the miners are trying to do the right thing.

有些矿工会试图破坏协议，或者行为不端。

Some of them may try to subvert the protocol, or they're behaving in a in a corrupt manner.

他们可能想发动双花攻击，或者其他让不同人看到区块链不同状态的攻击。

Maybe they were trying to create some sort of double spend attack or some attack that will basically cause different people to see different views on the blockchain.

而这个协议——也就是中本聪共识——能保证，即使有一部分矿工作恶，系统依然能提供一个统一、一致的账本视图。

And what this protocol basically does, Nakamoto Consensus, it's called, right, the Consensus Protocol of Bitcoin, is kind of guaranteed that even if a fraction of the miners are corrupt, this protocol still gives you a single consistent view of the of the ledger.

从某种意义上说，这和四十年前在状态机复制和拜占庭容错领域研究的问题完全是一回事。

And in a sense, that's exactly the thing that has been studied forty years ago, right, in state machine replication and Byzantine fault tolerance.

那么状态机复制，这和达成一致有什么区别呢？

And so state machine replication, so how does that compare to agreement?

从某种意义上说，核心问题其实就是对日志内容达成一致。

So in a sense, kind of the core problem is is just reaching an agreement on the content of the log.

现在你有一份日志，里面包含了一系列不同的指令。

So now you have a log that has different commands on them.

所以在这个情况下，日志就是区块链的一种抽象吗？

So a log would be an abstraction of a blockchain in this case?

是的。

Yeah.

你可以把区块链里的每个区块看作一组指令的集合。

So you could think about every block in your blockchain as just a set of commands.

然后这些指令连成一条链，这就构成了日志或者链。

And then you have kind of a, yeah, a chain of commands that would be kind of a log or a chain.

其实说的是一回事。

It it it's kind of the same thing.

但你真正想做的，不只是记录这些操作，而是要看执行它们之后会产生什么结果。

But really what you want to do is not just record things, but you want to try to see what is the outcome of executing them.

这时候就需要用到某种智能合约语言，或者像比特币脚本这样的脚本语言。

So this is where you had some sort of a smart contract language or or some scripting language like Bitcoin script.

在这种情况下，你可能需要验证一些签名，或者运行某个合约。

And in this case, you know, maybe you're you want to verify some signatures or or run some sort of a contract.

这就是整个过程中的执行环节。

So this is the execution part of it.

而这个非常强大的抽象概念，就是莱斯利·兰波特会提到的‘状态机复制’。

And so the very powerful abstraction, this is an abstraction that we'll hear about from Leslie Lamport, is what's called the state machine replication.

客户端其实是在和一个单一的状态机进行交互，就像它真的存在一样。

It's that clients are basically thinking as if they are interacting with a single state machine.

那什么是状态机呢？

So what is a state machine?

它就是一个系统，你给它发送指令，它就会根据这些指令，从一个状态切换到下一个状态。

It's simply a single system that you've send it commands, and through that command, it updates its state from one state to the other to the next.

每次执行一条命令，系统的状态就会更新到下一个状态。

And each time you kind of run a command, the state updates to the next state of the system.

在区块链的语境下，这个状态可能包括所有人持有的原生加密货币余额，还有本地存储和智能合约之类的。

I guess in blockchain context, that state would include, for example, everybody's balances in a native cryptocurrency plus, I guess, maybe like local storage and smart contracts.

这些都算是状态的一部分。

That would all be kind of part of this state.

对吧？

Right?

没错。

Exactly.

比如在比特币里，系统状态可能是我拥有一枚代币，如果我给你转账，状态就会变化——我的代币被作废，新生成一枚属于你公钥的代币。

So for example, in Bitcoin, maybe the state of the system is that I have a token, and maybe if I send a payment to you, then the way the the state would change is that now my token's erased, and there's a new token that forms that's under your public key.

对吧？

Right?

现在你就可以用这枚新代币了。

So now you can use that token.

所以这个状态实际上是在记录链上所有未花费的代币或交易。

So the state is kind of recording what are all the unspent tokens or transactions that are on the chain.

我觉得特别有意思的一点是，过去一段时间里，有两条并行的研究路线，至少在我看来，最近五年已经很大程度上融合在了一起。

So one thing that I think I find fascinating is, you know, you had for a while two parallel threads of research that at least it seems to me have converged to a large degree over the last five years.

我想听听你的看法。

I'm curious about your thoughts.

分布式计算的研究早就开始了，算起来已经有四十五年甚至更久，之后我们会听到兰伯特博士和利斯科夫博士谈到那些早期的工作。

You have distributed computing dates back again at this point you know forty five years or more and we'll hear from you know both doctors Lambert and Liskov about that early work.

而与此同时，比特币的出现带来了区块链技术，也催生了一个围绕它的研究社群，大家开始思考如何打造更完善的区块链协议。

And then separately, Bitcoin launched blockchain technology and sort of a research community around that, thinking about how to build better and better blockchain protocols.

比特币是2009年推出的，距离分布式计算最早的开创性工作，大概隔了三十年。

And that was in 2009, so that was maybe about thirty years after some of the earliest pioneering work on the distributed computing side.

所以除了时间上晚了三十年，我觉得这两条线在很长一段时间里其实是平行发展的。

And so in addition to being thirty years later, I felt like it was kind of a parallel thread for a while there.

但最近五年，我感觉这两条路正在慢慢汇合。

But now it seems to me that those two threads have been coming together over the past maybe five years or so.

所以，你是不是也这么看？

So I guess do do you sort of share that view?

如果是的话，你觉得最近是什么推动了这种融合呢？

And if so, like, do you think has been driving that sort of convergence recently?

是的。

Yeah.

当然了。

Absolutely.

首先，我觉得比特币确实是一场巨大的革命。

First of all, I really think that Bitcoin is just kind of huge revolution.

我们接下来可能会花很多时间聊这个。

We're probably gonna spend a lot of time talking about that.

但当初比特币出现的时候，人们其实并不清楚它到底在做什么。

But but in a way, when it happened, people didn't really understand what it's doing.

直到过了好几年，学术界和大家才逐渐明白，它其实解决了一个非常棘手的学术难题。

It actually took quite a few years for the research community and for everybody to understand that this is solving kind of a a very hard academic problem.

一旦大家明白比特币解决的是这个问题，就开始有人努力把它和经典理论联系起来。

Once it was clear that that is the problem that Bitcoin solves, there was this effort to try to connect that to this classical work.

你知道，这花了好几年时间。

You know, it took quite a few years.

大概在2016或2017年，情况才开始发生变化。

Around 2017 or '16 was when this changed.

我觉得最早的协议之一是Tendermint。

I think the first protocols, were Tendermint, as an example.

这些协议用的是经典的拜占庭容错算法，但把它应用到了权益证明机制上。

These were the protocols that kind of were using the classical Byzantine photonic protocols, but applying them on proof of stake.

所以，我认为真正关键的是从工作量证明转向权益证明，人们意识到，用权益证明也能实现比特币在工作量证明中达成的效果。

So, really, I think that this is this transition from a proof of work based protocols to proof of stake ones, and the realization that you can kind of mimic the same effect that Bitcoin obtained in proof of work using proof of stake protocols.

所以，我觉得2016到2017年这段时间，算是第一波浪潮。

So I think that was kind of the this first wave of 02/1617.

接下来值得一提的可能是Casper。

Maybe the next one to mention is Casper.

这是以太坊的最终性机制，大约出现在2017年。

This is the the Ethereum finality gadget, a period around 2017.

这个观点很到位。

That's a great point.

事实上，现在的理论已经清楚了：在工作量证明的环境下，传统的共识协议根本行不通，这跟某种形式的民间抵抗机制有着有趣的关联。

I mean, now even the theory is at the point where we understand there's a formal sense in which you kind of can't do the traditional consensus protocols in a proof of work context, which is an interesting interaction between kind of the type of civil resistance.

没错，如果你不知道参与者是谁，又没法做到一人一票，那就只能依靠某种稀缺资源来分配投票权。

Right, so if you don't know who people are and you can't do one vote per person because you don't know who the people are, so You have to do one vote per some scarce resource.

我所说的民间抵抗机制，指的就是这个。

That's what I mean by a civil resistance mechanism.

比特币就是靠工作量证明来实现这一点的。

And Bitcoin famously uses proof of work.

从某种角度说，这和兰珀特和莱斯科夫要讨论的那些技术是相互冲突的。

In some sense, that's incompatible with the types of techniques that doctors Lampert and Leskoff will be talking about.

而权益证明的民间抵抗机制，本身还有其他很多值得采用的理由。

Whereas proof of stake civil resistance, well there's a lot of other reasons you might want to use that as well.

你可能关心环保问题，或者可扩展性，诸如此类的原因。

Know maybe you're concerned about environmental reasons or scalability, what have you.

它其实也打开了拉姆波特和利斯科夫医生要介绍的那些技术的大门。

It also actually unlocks those techniques that doctors Lamport and Liskopf will tell us about.

这确实是个很好的观点。

So that's a great point.

当时这并不简单。

And it was not trivial at the time.

我记得2015年的时候，大家还在说，哦，有这个权益证明的想法，但到底该怎么实现呢？

So I I remember 2015, people were saying, oh, there's this idea of proof of stake, but how do we do it?

看起来根本不可能。

It seems impossible.

那时候很多人觉得，根本没法做出类似的东西。

You know, there were a lot of people that said that, you know, there there's no way to to kinda do do something similar.

所以那确实是一项了不起的突破，当时确实如此。

So it was a nontrivial advance, during those Absolutely.

比如，以太坊早在2015年上线主网之前，就已经在讨论权益证明了。

So for example, you know, Ethereum, know, was already talking about proof of stake before they even launched their original main net in 2015.

但真正转向权益证明，却是七年多以后，也就是2022年的事了。

And then the actual transition to proof of stake wasn't until 2022, right, seven seven, eight years later.

我同意你的看法。

So I I agree.

看来，实现这个方案的难度，远比最初提出想法要大得多。

It seems like that turned out to be a much more difficult problem than, you know, just the initial idea.

你提到了Tendermint，这个技术在Cosmos生态以及其他一些地方应该有不少人听说过。

So you mentioned Tendermint, which may be familiar to listeners from the Cosmos ecosystem among among other places.

你还提到了Casper，这是现在以太坊正在使用的机制。

You mentioned Casper, which is used in today's Ethereum.

现在的以太坊实际上有两层结构。

So today's Ethereum has sort of two layers.

底层是一个类似最长链的机制，上层则是一个基于Casper的最终性工具。

It has like a longest chain sort of lower layer and then a finality gadget based on top Casper.

正如你所说，Casper 中用到的技术，和我们今天嘉宾提到的那些方法关系非常密切。

And as you say, the techniques in Casper are gonna be very related to the techniques that that our guests tell us about today.

说实话，我个人觉得，最早让我注意到这种微妙互动的，是……

I actually, you know, on a personal note, I think I first became aware of this interesting interplay.

比特币协议一方面解决了学术界早已存在的问题，但它的解决方式却是前所未有的。

The idea that the Bitcoin protocol on the one hand was sort of solving a well established academic problem but then doing so in a way no one had ever thought about.

我想，最早指出这一点的，是你和达莉亚·马尔基合写的一篇综述。

Was a survey you wrote I think with Dalia Malki that that sort of first pointed that out.

那大概是 2017 年左右吧。

So that was maybe 2017 or something like that.

当时这件事给我留下了很深的印象。

That made a that made a big impression on me at the time.

是的。

Yeah.

事实上，要让整个社区普遍认识到拜占庭容错才是区块链的核心，花了好几年时间。

So it actually took, I think, quite a few years, for this kind of, prevailing, I guess, community understanding, right, that Byzantine fault tolerance is the core thing that that blockchains are doing.

但这一点其实并不明显。

But that actually wasn't obvious.

早期有一些协议做的并不是共识机制，也没解决拜占庭容错问题，直到后来才慢慢明朗起来，我觉得吧。

There were early protocols that kind of did things that were not consensus protocols or didn't solve the Byzantine agreement, and it took quite a while until kind of things settled, I I would say.

从2017年到2020年代，拜占庭容错领域的研究突然爆发了。

And and today, or at least, you even from 2017 to the twenty twenties, there was kind of an explosion of research in in Byzantine fault tolerance.

可以说，现在我们熟知的所有主流区块链，都在运行某种形式的拜占庭容错机制。

And and I would say, de facto, you know, all the major chains that we know are running some version of Byzantine thought tolerance.

区块链技术特别有意思的一点是，它为许多早已存在的计算机科学领域注入了新的活力，也带来了大量新资源。

One thing that's been wild, I'd say, about blockchain technology is it sort of breathed a lot of new life and also, frankly, a lot of new resources into a lot of areas of computer science that have been around for quite a while.

对吧？

Right?

另一个例子就是零知识证明的发展，几十年来，它一直被视为纯粹的理论构想。

So, like, another example would be say the development of snarks, which for many decades was viewed as a purely theoretical construct.

那时候大家觉得这玩意儿太玄乎了，根本不可能真正实现。

It was kind of, you know, something magical which you'd never hope to implement.

现在，我们真的看到高效零知识证明开始投入实际应用了。

And, you know, now we're really seeing very concretely efficient snarks come into production.

共识协议本来就是为了实用而设计的，我想今天我们会听到，区块链技术作为一项价值极高的应用场景，也让共识协议得到了极大推动。

And while consensus protocols, they were always sort of meant to be practical, I think we'll hear about today, they've been supercharged as well by having blockchain technology as a sort of extremely high value application of better consensus protocols.

但我还想跟你说个别的故事。

But let me even tell you another story.

那是2007年的事了。

So this was in 2007.

2007年，我参加了一个研讨会，目的就是探讨拜占庭容错到底能不能真正落地。

So 2007, I was at a workshop, and the goal of the workshop is to kind of see whether Byzantine fault tolerance is practical or not.

那时候，谷歌、雅虎和微软早就已经在用非拜占庭版本的一致性协议了，已经好几年了。

This was really a few years after Google and and Yahoo and Microsoft were using kind of the non Byzantine version of of agreement.

对吧？

Right?

那种协议属于Paxos一类的。

This is kind of Paxos type protocols.

所以问题根本不是共识协议能不能用。

So the question wasn't whether consensus protocols are practical.

问题是，你到底需不需要应对那种恶意的、不可预测的故障，而不仅仅是系统崩溃。

The question was, did you really need to be robust to potentially very adversarial or very unpredictable failures as opposed to just crashing.

是这样吗？

Is that right?

没错。

Right.

但当时其实有两个主要的质疑。

But there was actually two big complaints.

一个是说根本没人需要这个，另一个是说性能太差了。

One is that maybe nobody needs it, and the other was that the performance was horrible.

所以大家普遍觉得，这事儿不可能做得更好。

So people kind of didn't believe that it's possible to do it any better.

对吧？

Right?

他们说，这会慢得不得了。

They they said, oh, this is gonna be very slow.

我们现在有了很多高效的类Paxos算法。

We now have these very efficient Paxos like algorithms.

那些才是真正实用的。

Those are practical.

对吧？

Right?

很多云服务提供商都在用这些算法。

Those are used by a lot of these, cloud service providers.

但拜占庭容错？哦，那根本行不通。

But, yeah, Byzantine fault tolerance, oh, that's that's not practical.

那成本太高了。

That that's too expensive.

所以我觉得，这个领域确实经历了一段飞速发展的历程。

So yeah, I think we've seen this arc of a lot of work in a space kind of really exponentially improve it.

就说过去五年吧，区块链领域有没有出现什么让你觉得特别值得注意的共识机制创新？

Say like last five years, are there some sort of innovations in consensus but originating from the blockchain community that you find particularly notable?

现在不光是很多系统在用拜占庭容错，这个领域本身也涌现出不少新突破。

So we are seeing not just a lot of systems are using Byzantine fault lines, but we're also kind of seeing a lot of innovations in this space.

很多创新都集中在大幅提升吞吐量和降低延迟上。

A lot of it is focused on getting much higher throughput and much lower latency.

所以从用户角度看，高吞吐量意味着能容纳更多交易，而且交易费用也能更低。

And so I guess just to make that concrete, so from the user perspective, you want high throughput because you want there to be space for your transactions and even sort of you want it to be cheap to send your transactions.

低延迟就是你让区块链做点什么，比如转账，最好能几乎立刻完成。

A lot of latency just means whatever you ask the blockchain to do, like do a transfer or whatever you want it to happen ideally close to instantaneously.

这就是延迟的意思。

That would be the latency.

没错。

Yeah.

我们不妨回溯一下，从2009年的比特币，甚至到2017年早期的权益证明协议。

And maybe we can go back to Bitcoin 2009, or even 2017, the early proof of stake protocols.

那时候效率很低。

They were not very efficient.

比特币的话，每十分钟出一个块，或者几十秒一个块。

They had blocks every ten minutes if it's Bitcoin, or every tens of seconds.

吞吐量其实非常小。

And the throughput was actually quite small.

如果你想着要服务几十亿人，或者管理大型经济的系统，这种性能对很多应用场景来说根本不够。

And so, really, if you're thinking about serving, you know, billions of people or or or systems that really manage large economies, then that wouldn't be enough for some types of use cases.

所以你需要一个能服务大量用户的系统。

So you want to have something that can serve many, many users.

不仅如此，还得让用户感受到实时的体验。

And not only that, they can give them a real time experience.

服务大量用户，这就是高吞吐量。

So serving a lot of users, that's gonna be, you know, high throughput.

而让用户感觉实时，那就是低延迟。

And giving people kind of a real time experience, that would be kind of low latency.

一种创新是基于DAG的协议。

One type of innovation is these DAG based protocols.

这类协议有两层结构，能显著提升系统的吞吐量。

These are protocols that have two different layers, and they really push the throughput of these systems quite a bit.

比如SWE和MISTA SETI这类协议就体现了这一点。

We've seen this, for example, in SWE and protocols like MISTA SETI.

这算是一类重要的改进方向。

That's kind of one one family of major improvements.

另一类则专注于降低延迟。

And and the other one was a focus on trying to reduce latency.

能不能只用极少的通信轮次就达成拜占庭共识？

Can you kind of reach Byzantine agreement with very, very few route trips?

我们已经看到一些新协议采用了两种不同的运行模式。

And, you know, we've seen these new protocols that kind of have two different modes.

一种是常规模式，大概需要三次消息往返；还有一种快速路径，只需两次。

So a regular mode that maybe has three message delays, and a fast path that has just two message delays.

这已经是你能想到的最优方案了。

And this is the optimal thing that you could expect.

即使在非拜占庭的环境下，你也可以这么想。

Even in a non Byzantine setting, right, you could just think about this.

其实就是服务器把请求发给所有副本，等它们回复，交易就完成了。

It's really just this server sends it to all the replicas and gets a response back, and the transaction's committed.

所以，延迟已经降到理论上最低的水平了。

So, really, you get the smallest latency that you could possibly imagine.

对。

Yeah.

我觉得这种思路非常巧妙。

And I think this is a very cool approach.

这本质上是在优化最常见的场景，而这通常都是个好办法。

This is basically just optimizing the common case, right, which is just usually usually a good idea.

其实，这种设计在学术文献里早就有先例了，比区块链技术出现得还早。

And this itself has, you know, precursors in the academic literature that predate blockchain technology.

你知道，最出名的例子可能就是 AlpenGlow，这是索拉纳共识协议的新版本提案，我觉得它会在2026年推出。

You know maybe most famously with you know AlpenGlow which is the proposed new version of Solana's consensus protocol which I think will be rolled out in 2026.

这基本上算是这个想法的一个实际生产环境实现。大概有90%多的时间里，你可以跑得飞快，而在剩下那一点点时间里，性能损失微乎其微，甚至没有损失。

That's really kind of an in production implementation of this idea like maybe you know 90 some odd percent of the time, you can be super fast while losing little, if any, in the remaining percent from what you what you had before.

对。

Yeah.

我喜欢从系统设计角度来理解这一点，也就是说，你需要一种战时模式和一种平时模式。

The the way I like to think about this is from a systems design perspective, right, is that you wanna have kind of a a wartime mode and a peacetime mode.

所以在和平时期，也就是没有故障的时候。

So in peacetime, there's no failures.

一切都很顺利。

Everything is good.

这时候你就得追求极致的速度和效率。

You wanna be super fast and super efficient.

而这种情况其实占了绝大多数时间。

And that's gonna be most of the time.

这99%的时间你都能正常运行。

That's gonna be 99% of the time where you can work.

我们现在通过实际观察这些区块链的运行情况，已经能证实这一点了。

And we now we now empirically can be caught just watching all these blockchains in production.

从实际数据来看，大部分时间确实如此。

Empirically, we know that's most of the time.

没错。

Exactly.

这背后其实有加密经济的激励机制。

I mean, there's kind of crypto economic incentives.

对吧？

Right?

只要让攻击变得不划算，大家自然就不会去攻击了，以此类推。

So so once you make attacking it not efficient, then people will attack it less and so on.

关键是，你确实得能切换到战时模式。

And the thing is that you do wanna be able to switch to wartime.

对吧？

Right?

所以如果你正遭受攻击，你还是有办法应对，哪怕对方试图大规模破坏你的系统。

So if you are under attack, then you do have a way to kind of overcome, again, a massive attempt to corrupt your system.

我觉得这种双模式协议特别有意思，而且在这个世界里非常合理。

So I think these kind of dual mode protocols are fascinating and and I think make a lot of sense in this world.

最后再谈一下学术文献对现代区块链协议的影响。

One final thought on the the influence, you know, of the academic literature on consensus protocols to modern blockchain protocols.

我还想说，就连人们讨论区块链协议及其保障时所用的语言，也都深深植根于早期的研究基础。

I'd also say, like, even just the language people use to talk about blockchain protocols and the guarantees they have, I'd say is you know deeply, deeply informed by the foundations that were laid kind of on research side.

哪怕只是一个纯粹实用的区块链协议，人们依然期待它具备像‘部分同步下的最优容错性’这样的特性——这都是学术术语。

Like even if it's just meant to be a purely practical blockchain protocol, still is an expectation that you would have, for example, optimal fault tolerance in partial synchrony, which is a bunch of academic words.

这些概念最终都转化为对实际协议数学抽象的精确数学表述。

I mean it all translates to precise mathematical statements about sort of a mathematical abstraction of practical protocol.

但如今，这些已经几乎成了新一代区块链协议的标配。

But yet that has almost become sort of table stakes for new generations of blockchain protocols.

所以在我看来，人们对什么是好协议、什么是最前沿协议的整个思维方式，很大程度上受到了这个领域近半个世纪以来研究工作的深刻影响。

And so to me just the whole way people think about what makes a protocol good or sort of state of the art, that is I think very deeply shaped by the last, you know, now almost half century of work in this field.

确实如此。

Yeah.

对我们来说，这特别令人兴奋，因为我们原本是搞理论计算机科学的。

Mean, you know, for us, this is extremely exciting because we come from kind of a theoretical computer science.

而在这里，我们却面对着一个极其实用的系统。

And here, we have something that's extremely practical.

正如你所说，讨论这些协议、分析它们的方式，实际上正是借助了理论和数学抽象。

As you said, the language and the the way to think about these, to reason about these protocols, is actually using theory and mathematical abstraction.

这可不是个简单的想法，对吧？你有一个真实的系统，却用抽象的数学思维、用证明、用这套语言来理解它。

And this is a nontrivial idea, right, that you have a real world system, and you're actually using abstract mathematical thinking, and, you know, you're using proofs, and you're using that language in order to reason about your system.

对我来说更令人兴奋的是，这并不是什么新点子。

And what's to me even more exciting is that this is not a new idea.

正如我们待会儿会看到的，这种思路在分布式系统领域已经研究了四十年了。

This is an idea that, as we're gonna see, has been studied for forty years in in distributed systems.

所以，理论概念和数学模型与实际部署的系统之间存在着一种深刻而成功的联系，无论是二十年前的云计算，还是今天的几乎所有区块链协议，都是如此。

So there's kind of this very deep and successful connection between relatively theoretical notions and mathematical models and things that are just kind of in, I guess, in theory and practical systems that are deployed in real world, both in kind of, you know, clouds in the two thousands and today in in basically all blockchain protocols.

是的。

Yeah.

没错。

Right.

有句名言说，理论与实践之间的差距，在理论上总是比现实中要小。

The famous quote is that the gap between theory practice is always smaller in theory than in practice.

但我真的觉得，得益于研究和工程两方面许多聪明人的努力，这个差距正在不断缩小。

But I will say, I really feel like this gap has been getting narrower, thanks to the efforts of sort of lots of smart people both on the research side and on the sort of engineering side.

当然，这还没达到完全一致，但正如你所说，当前发展的理论确实旨在切实指导下一代生产级区块链协议的设计。

I mean, you know, it's not perfect convergence, but you know, like you say, I mean, really, the theory being developed right now really is intended to usefully inform the next generation of of in production blockchain protocols.

看到理论与实践之间这种协同效应，真的让人非常振奋。

It's just very exciting to sort of, you know, see that see that synergy between them.

确实如此。

Very much.

对我来说，这是因为这是一个双向的互动。

And and for me, I think it's because it's a two way street.

对吧？

Right?

并不是说有人躲在高塔里，整天写定理。

So it's not that, oh, there's somebody sitting in every tower and kind of, you know, writing theorems.

而是你看到实际中发生的情况，加以分析，然后发现：哦，原来这种双向的联系非常重要，而且对双方都带来了丰硕的成果。

It's actually you see what what's happening in practice, and you analyze that, and then you realize, oh, you can so so this two way connection is is extremely important and and was very fruitful, I think, for both both sides.

我想，无论是莉兹·兰波特还是芭芭拉·利斯科夫，任何人都能看出来，理论创新和实际解决方案之间存在着紧密的循环关系。

And, you know, I I would guess for both, you know, Liz Lamport and Barbara Liskov, right, I mean, there is you know, anyone can see there's a a really major loop between kind of the theoretical innovations and sort of the practical the practical solution.

相信我们之后还会从她们那里听到更多这方面的内容。

So I'm sure we'll hear more about that from them as well.

感谢收听本期a16z播客。

Thanks for listening to this episode of the a16z podcast.

如果你喜欢这期节目，别忘了点赞、评论、订阅，给我们打分或写评价，也推荐给你的朋友和家人。

If you like this episode, be sure to like, comment, subscribe, leave us a rating or a review, and share it with your friends and family.

想听更多节目，请前往YouTube、Apple Podcast和Spotify。

For more episodes, go to YouTube, Apple Podcast, and Spotify.

关注我们在X平台上的账号a16z，并订阅我们的Substack频道a16z.substack.com。

Follow us on x at a16z, and subscribe to our Substack at a16z.substack.com.

再次感谢收听，我们下一期再见。

Thanks again for listening, and I'll see you in the next episode.

请注意，本节目内容仅作信息参考，不构成法律、商业、税务或投资建议，也不应用于评估任何投资或证券，且不针对任何a16z基金的投资者或潜在投资者。

As a reminder, the content here is for informational purposes only, should not be taken as legal business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any a sixteen z fund.

请注意，a16z及其关联方可能仍持有本节目所讨论公司的投资。

Please note that a 16 and its affiliates may also maintain investments in the companies discussed in this podcast.

如需更多详情，包括我们的投资列表，请访问a16z.com/disclosures。

For more details, including a link to our investments, please see a 16z.com forward slash disclosures.