It was 2021. I was in Vancouver with my family, still in the throes of the COVID-19 pandemic. Restaurants closed, lockdown in place, I had nothing better to do than talk with @sreeramkannan. This is the beginning of Babylon. While the world felt like it was on pause, Sreeram and I were bouncing ideas off of each other. What would make a good startup? In this brainstorm, we got knee-deep into the idea of security sharing. Historically, it had been extremely hard to bootstrap a new blockchain’s security. When you launch, you need to get a massive amount of validators to secure the chain. It isn’t easy. So why not bootstrap security from an existing blockchain? This had actually already been written about back in 2010. Nakamoto suggested starting new chains - ones that wouldn’t clutter Bitcoin’s pristine nature, but could use its security. He called this merge mining. After researching more into this, Sreeram and I wondered if we can apply a similar model to the much newer Proof-of-Stake world. From this research, the first version of Babylon emerged: merged mining merged with Proof-of-Stake. Sreeram got busy with Eigenlayer, and @baby_fisherman and I continued to explore this concept. Very soon, we realize there is a significant flaw in merge mining: Bitcoin miners could be loyal to the Bitcoin blockchain and attack the newer one. That’s no good. We abandoned merge mining, but were still thinking about security sharing. Our second idea: Timestamping. You timestamp a Proof-of-Stake chain directly to Bitcoin itself, so that the miner can no longer attack the new chain alone. They’d have to attack Bitcoin to revert the timestamping information. This we’re happy with. We begin talking to many chains. We started working with the Cosmos ecosystem, thanks to the support of @zmania. We had been highlighting that this timestamping provides extra security that can solve the problem of long range attacks. It also speeds up the unbonding process, which used to take Proof-of-Stake chains 3 weeks. The crypto you’re trying to unbond could be a radically different price by that point. Voila, two big issues addressed. Yet, while many began easily integrating and enjoying our product, few were actually willing to monetarily invest in it. Why? We got some answers from the immensely brilliant @Minittowinit. Since our team was newer to crypto, we contacted Min to give us an outsider’s perspective on what we were building. After she performed her SWOT analysis, finding strengths, weaknesses, opportunities, and threats, she came back with the conclusion that we have too narrow a market because our protocol is a pure security product. How to broaden the market? One day we got an inspiration: what about the Bitcoin holders? After all, almost everyone in crypto has bitcoins. We start thinking about how to get the Bitcoin holders involved. At the time, 2 developments were already taking place: Eigenlayer restaking for autonomous verification systems (AVSs), and Mesh Security created by Sunny Aggarwal. So I met with @sunnya97. He asked: why not use Bitcoin to provide mass security? Instead of using a Cosmos blockchain asset, you use Bitcoin. Sunny then suggests we try to bridge the Bitcoin, and use it to give security to the chain. The security is entirely within the Cosmos ecosystem, but a new source of capital is being injected: Bitcoin. Great idea. The problem? There’s no known secure bridge from Bitcoin to another blockchain. This has been an issue for as long as Bitcoin has been around. While some have attempted creating a trustless bridge through a soft-fork upgrade called Drivechain, that effort has been stuck and we realize this may be too big of a hurdle for now. Instead, I insist that there must be a way to get the security without moving the Bitcoin. After all, Ethereum restaking and other remote staking of blockchains doesn’t involve moving the asset from the parent to the consumer chain. Why not mimic that design? Well, because Bitcoin doesn’t have smart contracts. That’s when we begin discussions with Orfeas Stefanos Thyfronitis Litos, a researcher at Imperial College London and member of @CommonPrefix. He mentions the concept of an extractable one time signature. This is the idea that if you do something “bad” on the Bitcoin blockchain – like sign a check twice sending the same money to different people – you get punished. Your private key gets revealed. We begin building our project around this concept. We used the two scripting functionalities of Bitcoin: timelocking, and spending by signing a private key. And after months of technical work, we finally came up with the design. A new native use case for Bitcoin: staking. No longer would Bitcoin just be a payment system. Through the collaboration and expansive efforts of all those mentioned in the tweet above, plus countless other unnamed contributors, Babylon was born as a way to provide Bitcoin security to smaller chains. Our story is one of pivots, commitment, and true teamwork as we persevered through every obstacle to end up with the final form of the startup we have today. Thank you to everyone who came with us on the journey.

Rollup-centric 路线图的合理性、权衡和思考以太坊估值的角度 受 @BanklessHQ @KyleSamani 播客启发，最近又回头思考了一下 @ethereum Rollup-centric 路线图。 1. Rollup-centric 路线图的合理性 @VitalikButerin 于 2020 年 10 月发表了「以Rollup 为中心的以太坊路线图」，提出以太坊需要在中短期内对 Rollup 进行集中支持。 其一，以太坊基础层扩容将聚焦于扩大区块的数据容量，而非提高链上计算或 IO 操作的效率。即：以太坊分片旨在为数据 blob（而非交易）提供更多空间，以太坊无需对这些数据进行解释，只确保数据可用。 其二，以太坊的基础设施进行调整以支持 Rollup（如 ENS 的 L2 支持、钱包的 L2 集成和跨 L2 资产转移）。长远来看，以太坊的未来应该作为安全性高的、人人可处理的单一执行分片，以及可扩展的数据可用性层。 此后 @VitalikButerin 在 2021 年 12 月发表的「Endgame」中描述了以太坊的最终图景：区块产出是中心化的，但区块验证实现去信任且高度去中心化，同时确保抗审查。 底层链为区块的数据可用性提供保证，而 Rollup 为区块的有效性提供保证。 类似于 Cosmos 的多链生态，以太坊的未来将是多Rollup 共存的——它们都基于以太坊提供的数据可用性和共享安全性。用户依赖桥在不同 Rollup 之间活动，而无需支付主链的高额费用。 上述两篇文章基本确定了以太坊的发展方向：优化以太坊的基础层建设，为 Rollup 服务。这些 Rollup 可以获得以太坊等效的安全性，并实现可以帮助以太坊实现超级扩展。 以上论点也许基于这样一个看法：既然 Rollup 已经被验证有效并且得到良好的采用，那么「与其将花上几年时间等待一个不确定且复杂的扩容方案（指原分片），不如将注意力放在基于 Rollup 的方案上」。 2. 合理性背后隐藏的权衡 L1 把执行下放到 L2，实际上隐含了两点权衡： - Gas 的收取以及 Gas 与ETH 的通缩关系。 首先，交易发生在 Rollup 意味着 gas 费由 Rollup 而非以太坊 L1 收取。其次，gas 实际上与 ETH 的通缩挂钩。在伦敦升级之后，EIP-1559 引入了燃烧机制，每笔交易的 base fee 会被燃烧。 在「The Merge」转向 PoS 之后，由于 ETH 发行的大幅减少和燃烧机制的存在，以太坊实际上进入了一个长达 20 个月的通缩周期。 根据 提供的公式 f(X) = X/108 （其中 X 为以太坊每日发行量，f(X) 为抵消所有新发行的 ETH 所需的 gwei/gas 价格），按撰写本文时 2,607 枚 ETH 日发行量计算，假设每个块的目标为 15x10^6 gas/block：gas 价格需要达到 24.1 gwei 才能抵消 ETH 的当日发行。 而由于过去几个月 L1 交易的低迷，L1 的 gas 价格基本上都低于这个数字。因此在谷底之后，ETH 回到了通胀趋势，总供应量有可能在不久后回到「The Merge」时候的水平。 如果未来更多的交易在 Rollup 发生，那么以太坊可能回到通胀。 - MEV 的价值捕获。 交易发生在 Rollup，意味着 MEV 由 Rollup 捕获（目前由团队运营的排序器控制），而非以太坊验证者获益。 当然，Rollup 向以太坊 L1 支付 DA、证明验证等「安全费用」，但普遍认为安全费用与执行层隐含的价值体量没有可比性。并且，在 EIP-4844 之后，DA 费用得到了进一步的降低。 3. L2 的现状是否与预期一致？ 我认为并不是与当初完全一致的。 从技术上讲，各个 Rollup 也许可以被看作是以太坊的「计算分片」。但从现实情况下，各个 L2 似乎存在一系列的问题亟需解决，例如 Vitalik 所提到的互操作性问题。 今天的 L2 数量太多，而且山头林立。ETH 虽然被作为主要的计价单位，但这个用例可能在之后被 L2 原生 token 所取代。更多的 Rollup 转向 Alt-DA 方案也可能削弱 L1 的盈利能力。总之，今天的 L2 不是劲往一处使，而是有相当激烈的竞争格局。 4. 以太坊应该怎样被估值？ @KyleSamani 提到由于比特币的特殊性，人们看待比特币和以太坊是不同的。以太坊有其完备性，我们倾向于认为从公司的角度来看待以太坊。 未来我们需要思考的问题是：怎样更好地进行以太坊的估值？未来几年，甚至十年后，以太坊应该在执行层的视角下还是安全层的视角下去估值和做定价？

After nearly four amazing years, I’ve decided to step away from @IOSGVC . It’s been the longest and most rewarding trek in my career, and I’m beyond grateful for every moment.