<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/">
    <channel>
        <title>sallygu.eth</title>
        <link>https://paragraph.com/@sallygu</link>
        <description>VC</description>
        <lastBuildDate>Thu, 23 Jul 2026 12:24:05 GMT</lastBuildDate>
        <docs>https://validator.w3.org/feed/docs/rss2.html</docs>
        <generator>https://github.com/jpmonette/feed</generator>
        <language>en</language>
        <image>
            <title>sallygu.eth</title>
            <url>https://storage.googleapis.com/papyrus_images/4a19d20758689fa994ee273704e75fdf92f738d914dd8134b7452a654517bc0b.png</url>
            <link>https://paragraph.com/@sallygu</link>
        </image>
        <copyright>All rights reserved</copyright>
        <item>
            <title><![CDATA[DA Layer Research]]></title>
            <link>https://paragraph.com/@sallygu/da-layer-research</link>
            <guid>slvFp6BLHA9OQy5COy6h</guid>
            <pubDate>Thu, 12 Oct 2023 14:41:37 GMT</pubDate>
            <description><![CDATA[DAC更decentralized的数据层mechanism of modular blockchain执行层：执行交易，状态更改，计算状态根结算层/共识层：验证zk/Optimistic rollup历史状态层：Portal Network存储数据的三方数据可用层：节点：全节点：安全，但需要存储、算力、带宽轻客户端：不能自己下载区块头，依靠临近全节点发送请求欺诈证明的漏洞：即只需要1个诚实的全节点监测区块的有效性，并在发现无效区块后构造一个欺诈证明、将其发送给轻客户端来提醒它们。但attacker可以通过隐藏部分区块数据来使欺诈证明无法生成，light client进而无法收到警告，或者恶意全节点发出假警告Reed-Solomon纠删码KZG多项式承诺（KZG Polynomial Commitments），用于证明多项式在特定位置的值与指定的数值一致，而无需包含该多项式的所有数据。Danksharding中通过采用KZG承诺来实现对纠删码的验证。单个区块的数据需要多个KZG承诺来保证。二维KZG承诺：降低节点负担从而避免中心化采用STARK和FRI的方案可用于证明纠删码的正确...]]></description>
            <content:encoded><![CDATA[<ul><li><p>DAC更decentralized的数据层</p></li><li><p>mechanism of modular blockchain</p><ul><li><p>执行层：执行交易，状态更改，计算状态根</p></li><li><p>结算层/共识层：验证zk/Optimistic rollup</p></li><li><p>历史状态层：Portal Network存储数据的三方</p></li><li><p>数据可用层：</p><ol><li><p>节点：</p></li></ol><ul><li><p>全节点：安全，但需要存储、算力、带宽</p></li><li><p>轻客户端：不能自己下载区块头，依靠临近全节点发送请求</p></li><li><p>欺诈证明的漏洞：</p><ul><li><p>即只需要1个诚实的全节点监测区块的有效性，并在发现无效区块后构造一个欺诈证明、将其发送给轻客户端来提醒它们。</p></li><li><p>但attacker可以通过隐藏部分区块数据来使欺诈证明无法生成，light client进而无法收到警告，或者恶意全节点发出假警告</p></li></ul></li><li><p>Reed-Solomon纠删码</p><ul><li><p>KZG多项式承诺（KZG Polynomial Commitments），用于证明多项式在特定位置的值与指定的数值一致，而无需包含该多项式的所有数据。Danksharding中通过采用KZG承诺来实现对纠删码的验证。单个区块的数据需要多个KZG承诺来保证。</p></li><li><p>二维KZG承诺：降低节点负担从而避免中心化</p></li><li><p>采用STARK和FRI的方案可用于证明纠删码的正确性</p></li></ul></li><li><p>Polygon Avail和Danksharding中采用了KZG承诺，而Celestia中采用了欺诈证明。两者各有优缺，KZG承诺无法抗量子，而欺诈证明依赖一定的诚实假设和同步假设</p></li></ul></li><li><p><strong>Rollup</strong></p></li><li><p>zk rollup</p><ul><li><p>DA: 任何人可以重建L2以抗审查</p></li></ul></li><li><p>optimistic rollup</p><ul><li><p>DA: L2所有数据被发布</p></li></ul></li><li><p>L2 cost = fixed + variable（gas + call data）数据可用支出占80-95%</p></li><li><p>解决da cost 问题：</p><ul><li><p>降低链上存储cost——eth搞proto-danksharding提供大区快</p></li><li><p>链下存储——通用: Celestia, Polyon Avail; Rolllup: StarkEx, zkPorter, Arbitum Nova; Validium: zk+off-chain</p><ul><li><p>StarkEx: 数据可用性委员会DAC维护的Validium方案</p></li><li><p>Arbitrum: AnyTrust DAC, Arbitrum Nova</p></li><li><p>zkPorter: toker holder(Guardians)质押维护，故障会罚没资金</p></li><li><p>都有Volition选项可以选链上或链下，安全性和成本的tradeoff</p><ul><li><p>DAC不太安全因为有勒索攻击，可以窃取委员会成员私钥威胁，Ronin Bridge和Harmony Horizon Bridge被盗</p></li></ul></li></ul></li></ul></li></ul></li></ul>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
        </item>
        <item>
            <title><![CDATA[MUSIC NFT Investment Thesis]]></title>
            <link>https://paragraph.com/@sallygu/music-nft-investment-thesis</link>
            <guid>yZZifV9hx1yUiKcp2oKe</guid>
            <pubDate>Thu, 12 Oct 2023 13:01:55 GMT</pubDate>
            <description><![CDATA[Web2 Music App OverviewNapster是最成功的音乐交换服务，它的出现使得P2P的音乐共享得以实现，用户可以互相share彼此的歌单，从而歌曲检索服务上实现了突破。1999年唱片公司联合上告Napster侵犯版权胜诉，标志在线音乐进入收费时代2002 Napster开创收费模式，用户可下载部分合作小唱片公司的音乐但仅能在电脑上播放，主流音乐无法搜到。用户于是通过错拼谐音等方式绕开版权检查。华纳音乐为代表的在线音乐服务倡导“为音乐付费”的新生活模式。Real推出RealOne（RealPlayer升级版）的网络视听中心，用户付费9.95美元可以从7.5万首歌曲中下载100首，主要音乐来自华纳、EMI、BMG、等大唱片公司，同时还可以接收NBA、MLB等现场直播和WSJ、CNN等新闻。Microsoft紧接着推出MSN音乐服务，选用PressPlay作为音乐内容提供商，收录索尼、环球、EMI等音源。同时提供个性化推荐选单，以及歌手动态、专辑、门票的全渠道通知等增值服务。HBO提供音乐及图像下载的同时，将票务和旅行折扣作为增值服务。MusicMatch提供精准推荐...]]></description>
            <content:encoded><![CDATA[<h1 id="h-web2-music-app-overview" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Web2 Music App Overview</h1><ul><li><p>Napster是最成功的音乐交换服务，它的出现使得P2P的音乐共享得以实现，用户可以互相share彼此的歌单，从而歌曲检索服务上实现了突破。</p></li><li><p>1999年唱片公司联合上告Napster侵犯版权胜诉，标志在线音乐进入收费时代</p></li><li><p>2002 Napster开创收费模式，用户可下载部分合作小唱片公司的音乐但仅能在电脑上播放，主流音乐无法搜到。用户于是通过错拼谐音等方式绕开版权检查。</p></li><li><p>华纳音乐为代表的在线音乐服务倡导“为音乐付费”的新生活模式。</p></li><li><p>Real推出RealOne（RealPlayer升级版）的网络视听中心，用户付费9.95美元可以从7.5万首歌曲中下载100首，主要音乐来自华纳、EMI、BMG、等大唱片公司，同时还可以接收NBA、MLB等现场直播和WSJ、CNN等新闻。</p></li><li><p>Microsoft紧接着推出MSN音乐服务，选用PressPlay作为音乐内容提供商，收录索尼、环球、EMI等音源。同时提供个性化推荐选单，以及歌手动态、专辑、门票的全渠道通知等增值服务。</p></li><li><p>HBO提供音乐及图像下载的同时，将票务和旅行折扣作为增值服务。</p></li><li><p>MusicMatch提供精准推荐服务，用户选定偏好的音乐流派和歌手后，自动生成符合偏好范围的歌单。相当于去除了广告，主持人，和不符合自己口味音乐的在线电台。</p></li><li><p>Napster被禁后，Morpheus填补了空白，用户电脑端的所有音乐和电影文件都可以被共享。其移除了Napster中心化存储的弊端，所有资料存储在用户个人电脑上，不设有单独的存储服务器。</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/a70fa96ec396e6f4073aca6c3c7127b44fb03fc13370ee3db84a85e16328a16a.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/81ee6366b9b2d92723f0bbcd31218c5c9410d53eb39eea0bfe1410c75a92b16a.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>音乐流媒体难盈利的另一个原因是音乐大盘太小，纯粹“音乐”这个媒介在视觉化时代完全无法和影视、体育、游戏竞争。疫情之前的 2019 年，美国录音音乐销售额是 $12b（包括流媒体、黑胶等所有非现场音乐销售），考虑到通胀是 1990 年代的一半。相比之下，有视觉部分的线下表演则好很多，2019 年美国收入 $55b，不过因为是纯线下所以规模化能力差。</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/5545a8512c0781a405c67dcb7fd2cd7527833dbddcc19b6850d75347a8cfa325.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h1 id="h-whats-the-problem-of-spotify" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">What’s the problem of Spotify?</h1><ul><li><p>Spotify的成功很大程度上建立在他分发渠道的多样和触达范围的广泛上。此外，他的使用门槛也很低，用户不订阅也能够使用app。</p></li><li><p>如今Spotify的订阅用户已经达到接近<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://www.nasdaq.com/articles/spotify%3A-still-not-convinced-about-its-profitability">200M</a>这也意味着他们发展到了一个相对成熟的阶段。Spotify的用户churn rate流失率非常低因为用户的歌单搭建和分享都长期依赖该平台，迁移成本很高。换句话讲，Spotify已经通过网络效应确保了用户粘性。</p></li><li><p>然而于此同时，Spotify也面对着new entrants like apple Music的冲击。在引流上的竞争使得赛道内的incumbents都需要用极高的广告支出来增加露出。同时唱片公司也不可能对流媒体分发平台保持忠诚而turn down更高的合作出价。</p></li><li><p>Spotify这类streaming平台的出现，主要<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://www.digitalmusicnews.com/2021/06/29/spotify-executive-entitled-pay-penny-per-stream/">解决的问题是对音乐的分发，而忽视了artists赚取收益的需求</a>。</p></li></ul><h1 id="h-whats-the-web3-solution" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">What’s the Web3 solution?</h1><ul><li><p>痛点：</p><ul><li><p>缺乏版权保护，这可能成为未来潜在的工业威胁。</p></li><li><p>版权的缺失极大地限制了 NFT 的价值。</p></li><li><p>版权可以将非物质文化遗产转化为传统意义上和金融意义上的真正资产。它通过版税、混音权和收听权使非物质文化遗产的收入来源多样化。</p></li></ul></li><li><p>NFT 解决方案：</p><ul><li><p>两种产业结构：</p><ul><li><p>大公司--唱片公司、出版商--音乐人</p></li><li><p>Spotify、YouTube、TikTok--音乐人</p></li><li><p>传统web2行业中音乐人处于极其不利和被动的地位</p><ul><li><p>缺少明确的版税收入</p></li><li><p>新音乐人的门槛</p></li></ul></li></ul></li><li><p>解决创作者公平收入回报问题：</p><ul><li><p>Web3音乐NFT可以通过NFT二级自由流转为独立音乐人提供更加鲁棒和垂直的收入来源</p></li><li><p>避免Spotify等第三方平台抽成而使价值无法完全回馈给真正的创作者</p></li></ul></li><li><p>解决数据确权问题：</p><ul><li><p>通过将音源等metadata打包上链，音乐数据和版权得到链上验证，具有更强的可溯源性</p></li></ul></li><li><p>解决粉丝互动问题：</p><ul><li><p>常见的方案也有将音乐NFT碎片化，使得散户/粉丝可以永久持有一小部分版权，这也对于构建长期具有粘性的粉丝经济有不错的助力</p></li></ul></li></ul></li><li><p>问题</p><ul><li><p>对于已有的web2平台和唱片公司来说，并不希望将既得的媒介利益返还给创作者或艺术家，架空自己的功能属性，在打市场的过程中必然遭到巨大的阻力</p></li><li><p>由其对于头部的唱片发行公司而言，一个顶级歌手所能带来的收入和关注度是无法被简单计算和替代的，这也解释了为何在音乐NFT平台叙事炙手可热的2021年，也并没有成功找到PMF实现破圈的产品案例</p></li><li><p>BD-heavy 同时也面临 adoption问题 很可能会是一个winner-take-all的行业</p></li></ul></li></ul>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
        </item>
        <item>
            <title><![CDATA[Fan Token Research]]></title>
            <link>https://paragraph.com/@sallygu/fan-token-research</link>
            <guid>q7o5uJeodUwGz1aUOr5i</guid>
            <pubDate>Thu, 12 Oct 2023 11:56:07 GMT</pubDate>
            <description><![CDATA[BackgroundCooperation between football clubs and cryptocurrencies has increased rapidly since the revenues of clubs have fallen sharply under the COVID-19, with limited gains from traditional methods. European clubs reportedly haemorrhaged €1.1bn in total. As a result, they began to seek out moneyed investors, with Saudi Arabia’s purchase of Newcastle FC being one recent example. Over the past few years, clubs have also embraced support from the newly wealthy cryptocurrency industry, which ha...]]></description>
            <content:encoded><![CDATA[<h2 id="h-background" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Background</h2><p>Cooperation between football clubs and cryptocurrencies has increased rapidly since the revenues of clubs have fallen sharply under the COVID-19, with limited gains from traditional methods. European clubs reportedly haemorrhaged €1.1bn in total. As a result, they began to seek out moneyed investors, with Saudi Arabia’s purchase of Newcastle FC being one recent example. Over the past few years, clubs have also embraced support from the newly wealthy cryptocurrency industry, which has quickly cottoned on to the enormous revenue potential of football’s online fandom.</p><h2 id="h-advantages" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Advantages</h2><ul><li><p>Financing without affecting the shareholding structure</p></li><li><p>Expand overseas presence without geographical restrictions</p></li><li><p>Increase the sense of involvement of fans in the joint governance of the club</p></li></ul><h2 id="h-operation-model" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Operation model</h2><p>Take the example of Socios&apos; partnership with several clubs.</p><p>Fan engagement platform <a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="http://Socios.com">Socios.com</a> and blockchain company Chiliz work together to develop fan tokens, which are initially sold at a fixed price and subsequently traded into the marketplace. If the initial sale reaches a minimum, Socios takes a profit from the subsequent transaction fees and a cut of each token sale; around 50%. The rest goes to the club (in tokens or cash).</p><p>Fans are given a voice in the governance of the club based on the percentage of coins held, such as jersey design, logo design, how wins are celebrated, etc.</p><h2 id="h-clubs-involved" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Clubs involved</h2><p>Roma, Paris Saint-Germain, Inter Milan, AC Milan, Lazio, Manchester City, Atletico Madrid, Juventus, Argentina national team, Spaniards ......</p><h2 id="h-risks" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Risks</h2><ul><li><p>Fan tokens themselves are not very liquid but can be passively involved in the dramatic fluctuations of the cryptocurrency market.</p></li><li><p>Market risk is passed on by clubs to ignorant fans, especially many young people.</p></li><li><p>One such partnership, Lazio&apos;s partnership with Binance, to issue tokens directly on the CEX platform, is seen as encouraging fans to participate in highly leveraged transactions.</p></li></ul>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
        </item>
        <item>
            <title><![CDATA[Sandbox Case Study]]></title>
            <link>https://paragraph.com/@sallygu/sandbox-case-study</link>
            <guid>iz4qBK5OTSD4cO36AR6S</guid>
            <pubDate>Thu, 12 Oct 2023 11:53:06 GMT</pubDate>
            <description><![CDATA[IntroductionSandbox 是一个基于以太坊区块链的去中心化虚拟游戏世界。创作者可以创作3D像素化资产，并将其以NFT 的形式导入游戏中，在未来The Sandbox可能会把真实世界的物体扫描导入到游戏中。The Sandbox 属于典型的 Play-to-earn（P2E）模式 —— 与传统游戏不同，玩家或艺术工作者、游戏设计师可以通过 The Sandbox 提供的免费软件，例如 VoxEdit 和 Game Maker 为自己打造数字资产和应用，例如艺术画廊、3D模型甚至游戏。所有被打造出来的数字资产都能与他人分享或出售，从而获得利益。Data最新迭代的sandbox拥有超过120万MAU，高峰期可达260万。已经创建了70,000,000个世界；每天有超过100,000个世界被创建。被BlockchainGamer.biz评为2019/2020年最值得期待的区块链游戏Mechanism1.LANDLAND以以太坊上不可伪造的ERC-721代币的形式代表了沙盒元域中的数字房地产。在Sandbox地图上会显示所有的LAND，游戏以SAND的价格将这些LAND、出售...]]></description>
            <content:encoded><![CDATA[<h1 id="h-introduction" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>Introduction</strong></h1><ul><li><p>Sandbox 是一个基于以太坊区块链的去中心化虚拟游戏世界。</p></li><li><p>创作者可以创作3D像素化资产，并将其以NFT 的形式导入游戏中，在未来The Sandbox可能会把真实世界的物体扫描导入到游戏中。</p></li><li><p>The Sandbox 属于典型的 Play-to-earn（P2E）模式 —— 与传统游戏不同，玩家或艺术工作者、游戏设计师可以通过 The Sandbox 提供的免费软件，例如 VoxEdit 和 Game Maker 为自己打造数字资产和应用，例如艺术画廊、3D模型甚至游戏。</p></li><li><p>所有被打造出来的数字资产都能与他人分享或出售，从而获得利益。</p></li></ul><h2 id="h-data" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>Data</strong></h2><ul><li><p>最新迭代的sandbox拥有超过120万MAU，高峰期可达260万。</p></li><li><p>已经创建了70,000,000个世界；每天有超过100,000个世界被创建。</p></li><li><p>被BlockchainGamer.biz评为2019/2020年最值得期待的区块链游戏</p></li></ul><h2 id="h-mechanism" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Mechanism</h2><h3 id="h-1land" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0"><strong>1.LAND</strong></h3><ul><li><p>LAND以以太坊上不可伪造的ERC-721代币的形式代表了沙盒元域中的数字房地产。</p></li><li><p>在Sandbox地图上会显示所有的LAND，游戏以SAND的价格将这些LAND、出售给用户。</p></li><li><p>一个LAND代表一个服务器，所有者可以决定运行的游戏、实施的游戏机制、使用的资产，以及玩家是否可以加入（以及他们是否必须购买游戏或免费玩）。</p></li><li><p>Sandbox元宇宙由166,464个LAND组成，这些土地可以拼凑起来形成ESTATES。</p></li></ul><h3 id="h-2assets" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0"><strong>2.ASSETS</strong></h3><ul><li><p>ASSETs是由用户创建的ERC-1155代币，可以用来填充LANDs。</p></li><li><p>创建ASSETs后，玩家可以自由定制并在游戏中使用它们，移动到其他世界，或与平台上的用户交易。</p></li><li><p>可以在市场上购买 ASSETs，玩家不一定要创造自己的游戏道具。</p></li></ul><h3 id="h-3catalyst" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0"><strong>3.Catalyst</strong></h3><ul><li><p>Catalyst是在使用时被销毁的ERC-20代币，用来定义ASSET的等级和稀缺性。Sandbox元宇宙中最强大的ASSET是具有最高等级和最低稀缺性的道具。</p></li><li><p>通过在道具上添加Catalyst，NFT将获得一个空口袋，可以用宝石填充。Catalyst越好，该道具就有越多的口袋</p></li><li><p>获得Catalyst的唯一途径是通过质押Sandbox的SAND代币。</p></li></ul><h3 id="h-4gem" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0"><strong>4.Gem</strong></h3><ul><li><p>Gem也是使用时会被销毁的ERC-20代币，玩家可以用它来定制他们的ASSETs</p></li><li><p>将Gem放入ASSET的口袋中，以定义它们的属性和主要属性（如25点力量和50点魔法）。</p></li><li><p>道具有多大的用处取决于你所填充的Gem。</p></li><li><p>可以通过质押SAND来获得Gem。</p></li></ul><h3 id="h-5" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0"><strong>5.头像</strong></h3><ul><li><p>头像是玩家在游戏中的一个体素代表，可以根据个人喜好进行定制。</p></li><li><p>可以使用资产来为他们的头像配备衣服、盔甲、武器和其他游戏中的道具。</p></li></ul><h2 id="h-tokenomics" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>Tokenomics</strong></h2><ul><li><p>Sandbox 的平台代币是 SAND，是一种 ERC20 代币，总供应量30亿枚，目前流通量为7.05亿枚（流通率23%左右）。</p></li><li><p>SAND 在 Sandbox 元宇宙经济系统中的地位极高，主要用于线上交易、治理和质押</p></li><li><p>所有以 SAND 代币进行的交易额的 5%(交易费用)将分配 50%给质押池，作为质押SAND 代币的代币持有者的奖励，50%给“基金会”；基金会提供资助来激励平台上的高质量内容和游戏制作。</p></li></ul><h2 id="h-revenue-streams" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>Revenue Streams</strong></h2><ol><li><p>LAND销售</p></li><li><p>公司NFT销售</p></li><li><p>玩家订阅&amp;服务</p></li></ol><p>25%staking pool</p><ol><li><p>交易费</p></li></ol><p>5%SAND交易转账费用，50%staking pool, 50%Foundation</p><ol><li><p>广告收益</p></li></ol><p>25%staking pool；75%Foundation</p><h1 id="h-team" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>Team</strong></h1><ul><li><p>The Sandbox是沙盒系列游戏的最新版本。由Sebastian Borget和Arthur Madrid于 2012 年创立，由位于旧金山的软件公司Pixowl制作。</p></li><li><p>继 2017 年Pixowl 宣布计划将 Sandbox 转变为基于以太坊的 3D 游戏，其中用户创建的每个游戏内物品都是NFT。</p></li><li><p>2018 年，Pixowl 被总部位于软件公司Animoca Brands收购。</p></li><li><p>Sandbox 在 2019 年和 2020 年通过三笔 SAND 代币销售筹集了约 700 万美元，自 2019 年以来已从各种土地销售中筹集了数百万美元。</p></li></ul><h1 id="h-funding" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>Funding</strong></h1><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/eee4df63baf4bef201ef9f8833c8bfdc2f48d83bae8b515654bfdb2759862b53.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
        </item>
        <item>
            <title><![CDATA[How to make your DAO alive]]></title>
            <link>https://paragraph.com/@sallygu/how-to-make-your-dao-alive</link>
            <guid>hwnuKKsNPUsBTQqywdOq</guid>
            <pubDate>Thu, 12 Oct 2023 11:40:10 GMT</pubDate>
            <description><![CDATA[Building cohesionPhilosophy作为社会性动物，人渴望被承认，也就是别人的目光。但是同时，当别人的目光围拢过来的时候，他又感到窒息，感到不自由。 一个优秀的社区应该要寻找到渴望承认与自由的平衡点。而一切倡议的成功秘诀，不过在于用一种逻辑自洽的方式，帮助人们减缓肩负自由的疲惫与孤独。(refer to弗洛姆《逃避自由》)Solutions一起发掘项目投项目，就像IC meeting一样。投项目这个动作的本质是voting，voting makes people feel they are involved, seen and recognized一起上课学习Web3听AMA，就像回到大学时光。强迫性社交其实能让你避免孤魂野鬼的生活，which is 社会主义大家庭的优越性分享code和对web3的见解，满足炫耀欲和表达欲，知乎的产品逻辑吐槽web2工作职场分享八卦，拥有共同的回忆和敌人可以增强团体认同陌生人社交，networking，人人网领英那套共同建设DAO，领取任务获取激励，这种激励可以是一种身份名誉性的激励，let‘s say发放NFT/POAP，也可...]]></description>
            <content:encoded><![CDATA[<h2 id="h-building-cohesion" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Building cohesion</h2><h3 id="h-philosophy" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">Philosophy</h3><p>作为社会性动物，人渴望被承认，也就是别人的目光。但是同时，当别人的目光围拢过来的时候，他又感到窒息，感到不自由。 一个优秀的社区应该要寻找到渴望承认与自由的平衡点。而一切倡议的成功秘诀，不过在于用一种逻辑自洽的方式，帮助人们减缓肩负自由的疲惫与孤独。(refer to弗洛姆《逃避自由》)</p><h3 id="h-solutions" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">Solutions</h3><ul><li><p>一起发掘项目投项目，就像IC meeting一样。投项目这个动作的本质是voting，voting makes people feel they are involved, seen and recognized</p></li><li><p>一起上课学习Web3听AMA，就像回到大学时光。强迫性社交其实能让你避免孤魂野鬼的生活，which is 社会主义大家庭的优越性</p></li><li><p>分享code和对web3的见解，满足炫耀欲和表达欲，知乎的产品逻辑</p></li><li><p>吐槽web2工作职场分享八卦，拥有共同的回忆和敌人可以增强团体认同</p></li><li><p>陌生人社交，networking，人人网领英那套</p></li><li><p>共同建设DAO，领取任务获取激励，这种激励可以是一种身份名誉性的激励，let‘s say发放NFT/POAP，也可以是直接的经济激励，talking about token, grant</p></li></ul><h1 id="h-building-sustainability" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Building sustainability</h1><h3 id="h-philosophy" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">Philosophy</h3><p>在沉没成本足够大时，人们无法轻易撤离。这是因为人作为非理性动物，常常会赋予自己耗费的精力和时间过多毫无必要的意义。尤其是对于心智化程度较低的年轻一代，由于缺乏对自我和外界的深度校准，他们容易对自己熟悉的观念与事物，产生一种近乎cult般的盲从，而忽视潜在的危机。从1929年的大萧条，到2008年次贷危机，再到如今层出不穷的庞氏，历史的重演完美地印证着这种人性牢笼的经久不衰。</p><h3 id="h-solutions" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">Solutions</h3><p>staking其实就是这样的思路</p><p>所以可持续性实质上是建立在人的惰性和惯性上，再深一点来讲，这种惰性和惯性是人对未知的恐惧和不安全感。无法对抗这种求生意识的人，不可能做到及时止损。大部分人放弃一件事，通常也只是因为边际效用递减和缺乏毅力。在一个具备合理激励梯度的</p><p>设立具有周期性的活动，培养成员的参与惯性</p>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
        </item>
        <item>
            <title><![CDATA[MEV Research]]></title>
            <link>https://paragraph.com/@sallygu/mev-research</link>
            <guid>cwmoMGIfxdrYGfgNrR6Q</guid>
            <pubDate>Thu, 12 Oct 2023 01:48:56 GMT</pubDate>
            <description><![CDATA[概念MEV即通过在区块中添加和排除交易并更改区块中的交易顺序，从区块生产中提取的超过标准区块奖励和燃料费用的最大值在以太坊中一个交易发起后，这笔交易会被放在 mempool（一个保存待执行交易的池子）中等待被矿工打包矿工就可以看到 mempool 中的所有交易，而矿工的权利是很大的，矿工掌握了交易的包含、排除和顺序如果有人通过支付更多的 Gas 费贿赂矿工调整了交易池中的交易顺序而获利，这就属于一种最大可提取价值 MEV目前有两种MEV形式最为常见：一种为Arbitrage套利，另一种为三明治攻击 （liquidation排在第三）以三明治攻击为例，其可以通过在链上监控大额的 DEX 交易抢跑大额订单随后卖出比如有巨鲸想在 Uniswap 上购买价值 100 万美金的山寨币，而这一笔交易会将这个山寨币的价格拉高很多这笔交易被放入 mempool 的时候，监控机器人发现有机会，就贿赂打包这个区块的矿工将一笔买入操作插队在这个人前面，然后在巨鲸购买操作后插入卖出操作，获取交易拉盘利润尽管JIT(Just-in-time Compilations)和清算也算MEV策略，但其volume...]]></description>
            <content:encoded><![CDATA[<h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">概念</h2><ul><li><p>MEV即通过在区块中添加和排除交易并更改区块中的交易顺序，从区块生产中提取的超过标准区块奖励和燃料费用的最大值</p><ul><li><p>在以太坊中一个交易发起后，这笔交易会被放在 mempool（一个保存待执行交易的池子）中等待被矿工打包</p></li><li><p>矿工就可以看到 mempool 中的所有交易，而矿工的权利是很大的，矿工掌握了交易的包含、排除和顺序</p></li><li><p>如果有人通过支付更多的 Gas 费贿赂矿工调整了交易池中的交易顺序而获利，这就属于一种最大可提取价值 MEV</p></li></ul></li><li><p>目前有两种MEV形式最为常见：</p><ul><li><p>一种为Arbitrage套利，另一种为三明治攻击 （liquidation排在第三）</p></li><li><p>以三明治攻击为例，其可以通过在链上监控大额的 DEX 交易抢跑大额订单随后卖出</p></li><li><p>比如有巨鲸想在 Uniswap 上购买价值 100 万美金的山寨币，而这一笔交易会将这个山寨币的价格拉高很多</p></li><li><p>这笔交易被放入 mempool 的时候，监控机器人发现有机会，就贿赂打包这个区块的矿工将一笔买入操作插队在这个人前面，然后在巨鲸购买操作后插入卖出操作，获取交易拉盘利润</p></li></ul></li><li><p>尽管JIT(Just-in-time Compilations)和清算也算MEV策略，但其volume相对来说还比较小</p><ul><li><p>更高级的策略还包括杠杆三明治交易、Just-in-time bot、流动性再平衡等</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/037f14766a5e3beb2d6925c7a5af7e3c9f992763af1244867a315fbab6ad7e31.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>简而言之，整体MEV的供应链流程如下</p><ul><li><p>用户发送意向交易到钱包</p></li><li><p>钱包推送意向交易到memopool，最后送到searcher手中</p></li><li><p>searcher在memopool中进行搜索后，发送bundle给builder</p></li><li><p>builder将MEV和其他交易打包到区块内</p></li><li><p>在merge之后，builder还需要聚合交易并交给验证者验证</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/248c1c0b3cf5e4fb0df45375f95b649cf845bb55c9198b742109cc56ac976f79.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">背景</h2><h3 id="h-pbsproposer-builder-separation" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">PBS（Proposer-Builder Separation）</h3><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/56d9035b588be57893ab5dde3874b076e975effe73eedfaa6f4c21212c0f2236.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>由于darksharding1.0的缺陷无法解决MEV问题，以太坊坎昆升级后在技术路线图规划上，提出了提议者-打包者分离（PBS）</p><ul><li><p>指在通过MEV外部化的思路，使整个MEV过程更加流程正义</p></li><li><p>预计落地需要1-3年时间</p></li><li><p>主要是解决节点的分工问题</p></li></ul></li><li><p>高性能节点作为builder就下载Blob数据编码分发，然后创建blcok广播给低性能节点抽查</p><ul><li><p>全节点</p></li><li><p>同步数据量和带宽要求较高，相对中心化</p></li></ul></li><li><p>低性能节点就作为Proposer负责抽查检验数据有效性，然后创建和广播blockheader</p><ul><li><p>轻节点</p></li><li><p>同步数据量和带宽要求较低，去中心化一点</p></li></ul></li><li><p>高性能节点作为builder肯定权利更大，有可能会故意不打包一些交易或者随便编序插入自己想插入的交易去捕获mev机会</p></li><li><p>所以为了防止权力滥用，低性能节点propoesr就会公布一个抗审查清单（crlist），把所有mempool所有交易都装进去</p><ul><li><p>这样一来builder不能随便把mempool里的交易扔掉除非有gas限制，或者插入不属于这个mempool的自己的交易搞MEV套利</p></li><li><p>不过builder仍然有权力对mempool里的交易进行排序，还是有一定套利空间</p></li></ul></li><li><p>在打包完之后，builder就会把交易列表hash广播给低性能节点proposer，由其自己选一个交易列表生成blockheader广播</p></li><li><p>节点同步数据就是一方面从proposer那里要blockheader，一方面问builder要block的body</p></li></ul><h3 id="h-pbstwo-slot-proposer-builder-separation" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">双槽PBS（Two-slot Proposer-Builder Separation）</h3><ul><li><p>主要是为了解决builder捕获MEV几家独大的中心化问题</p></li><li><p>模式：builder卷起来对crList进行竞价</p><ul><li><p>主要就是不同的builder拿到crlist排除交易列表和header之后出价竞拍</p></li><li><p>proposer就选择竞价成功的header和builder，拿中标费</p></li><li><p>Committees确认中标的header后Builder披露中标的block body</p></li></ul></li><li><p>虽然仍然有排序的MEV套利空间，但是builder会因为要竞争出块而互相内卷价格，是的MEV行为的利润收窄</p><ul><li><p>但因为高性能节点其实数量有限，builder本来就少，有MEV经验的builder可能还是会吃掉大部分竞标，最终还是相对中心化</p></li></ul></li><li><p>这个机制比较大的问题还在于双槽（Two-slot），有两个插槽 slot的设计</p><ul><li><p>对比原有方案，有效的出块时间会被延长 24 秒（一个 slot=12 秒）</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/35a30792a9b392923c7c93894a4ea77c13afa4d89b21c49adaefe952c1e34d4c.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h3 id="h-epbsenshrined-proposer-builder-separation" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">ePBS（enshrined Proposer-Builder Separation）</h3><ul><li><p>封装提议者与构建者分离ePBS相当于是进阶版的PBS</p><ul><li><p>指直接使用协议功能构建的PBS，有更强的信任假设</p></li><li><p>但完成度相对比较低，需要的周期更长</p></li><li><p>其他渐进版本PEPC和OptimisticRelaying也在引起越来越多关注</p></li></ul></li><li><p>封装的私有内存池可以在一定程度上保护应用dapp用户遭到抢跑攻击</p><ul><li><p>应用dapp内交易公开所有人可见的问题是容易被bot发现抢先交易</p></li><li><p>所以创建“私有/加密内存池”，用户的交易在被打包撞到一个区块里之前都是被加密的</p></li><li><p>但这对加密技术有所要求，一个是用户不能率先涌入解码，一个是交易被不可逆地接受后其可以自动解码</p></li></ul></li><li><p>目前对中心化运营商进行加密的几种模式</p><ul><li><p>时间锁加密：按照一定顺序计算任何人都能解开，且无法并行计算</p></li><li><p>阈值加密：由一个可信的多数committee解密数据，类似enshrined信标链</p></li><li><p>可信硬件：SGX</p></li></ul></li><li><p>不过尽管功能很有价值，但目前这些加密方式都还没有建立起充分的信任基石</p><ul><li><p>在延迟加密或一些其他技术有所突破前，L1都不太可能接受</p></li><li><p>此外，封装太多内容可能也会使协议过于复杂，增加治理负担</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/a09eb120ae76ebf5682c31426c8237280dcbd714cd66e5ea5a1fb52b2000595c.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">市场数据</h2><ul><li><p><a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://dune.com/chorus_one/ethereum-mev-data">https://dune.com/chorus_one/ethereum-mev-data</a></p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/47da2c1eae94ed35c4005ee6eb150ab0fc8424de509e5848ca00dc83b9844a66.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>Flashbot社区发布的5&amp;6月报显示在合并接近一年后，MEV-Boost的relay和builder市场都进入了稳定期。目前Flashbot、Ultra Sound、Agnostic、Bloxroute等4家relay提供商占据了总市场份额的 80%，同样builder市场的大部分市场份额也由头部5家占据。</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/4bffedf48d31823baf8448a6a097c647c7961a4752449422297deddf61e53edc.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/ed74210df3c3d552b19db0d29a3f8f9ae188bae8b92c802ae8b821a2d85f3ac3.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">机制分类</h2><h3 id="h-mev-auction" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">MEV Auction</h3><ul><li><p>Merge前PoW的玩法</p></li><li><p>主要参与方：searcher和矿工</p><ul><li><p>矿工即builder的角色，有出矿权+可以通过排序、插队和篡改等捕获MEV价值</p></li><li><p>searcher内卷gas费确保自己的交易被先打包，内卷严重可能会拿出90%MEV利润出让</p></li><li><p>在这种机制上矿工占很大优势地位，同时交易隐私和安全也不太好</p></li></ul></li><li><p>Flashbots之类的项目方有做searcher和矿工之间的私密通信渠道，防止恶性竞争本质上是增强信任度</p><ul><li><p>searcher先打包交易成bundles提交到auction系统里，设个限价最低价</p></li><li><p>矿工卷竞价抢自己感兴趣的bundle（像PBS之后抢crlist一样），也设个限价心理阈值</p></li><li><p>矿工从auction中选择竞价最高最能获利的 Bundle包含到block里，block里交易和 Bundle 进而就按照讲好的顺序和价格执行&amp;结算</p></li><li><p>这一套流程好处是互相卷，双向选择</p></li></ul></li></ul><h3 id="h-mev-boost" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">MEV-Boost</h3><ul><li><p>Merge之后PoS时代，MEV-Boost 取代auction成为主流</p><ul><li><p>矿工被validator/节点替代（在信标链质押 32 个 ETH +跑以太坊客户端即能成为validator）</p></li><li><p>PBS之后，validator一部分成为builder一部分成为proposer，proposer从builder那里获得区块</p></li><li><p>减轻了MEV对以太坊的中心化压力也推动validator内卷</p></li><li><p>90% 的validator在第二个月底前运行了 MEV-Boost</p></li></ul></li><li><p>加入了relay的角色，负责验证区块和托管区块</p><ul><li><p>任何人都可以跑一个relay，builder可以连接连 7 个不同的中继器</p></li><li><p>这些relay也可以收到27 个不同builder的竞标</p></li><li><p>相比较builder的市场，relay市场还是较为去中心化</p></li><li><p>目前Ultra Sound Money占据三分之一的市场份额，Bloxroute、flashbots和agnostic均占19%左右</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/fc6928ac4490d5d9dc5434ccc650cf3844ec52b3a3e5b4f62148777b37a35e85.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>整套工作流PoS validator 必须跑三个软件：</p><ul><li><p>验证客户端、共识客户端和执行客户端</p></li><li><p>MEV-boost 相当于是信标节点的辅助开源软件，用于查询将block building外包出去的builder网络</p></li><li><p>Block builder准备完整的区块，把MEV 提取和奖励机制优化之后就把block提交给relay</p></li><li><p>就上面说的，relay可以汇总多个builder的区块，然后挑一个最赚钱的区块</p></li><li><p>同样builder也是通过MEV-boost 实例配置连多个relay，从共识层客户端挑个最赚钱区块提交打包给主网</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/229fe83923796613359f55fc8110ddc6da5831ec11a4b9637b728c16ea38f8c0.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>MEV-boost的机制也为proposer创收</p><ul><li><p>merge前propser在 MEV 收入中的份额平均低于 50%。</p></li><li><p>今年起越来愈多builder涌入竞标，转给proposer的MEV利润分红就逐渐稳定在 50% 以上</p></li><li><p>整体走势上看propser获得原子 MEV 利润更在攀升，最高能达到夸张的95%</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/2bcb2b60714c00c067d6cb332ad5b9ae7c5808e723cd6a512ab6271c9d408e1c.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h3 id="h-suave" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">SUAVE</h3><p><strong>机制阐释</strong></p><ul><li><p>虽然 MEV-Boost 有带来种种好处，但也还是面临builder中心化的问题，Suave意在实现中心化民主化的MEV供应链（建设者、中继器、集中式 RFQ 路由等）生态</p><ul><li><p>允许开发者在solidity环境下搭建应用</p></li><li><p>帮他们在链下跑这些应用并采用TEE确保隐私 [图片]</p></li></ul></li><li><p>可以简单地把SUAVE的mempool想象成一个高德/Uber叫车市场，</p><ul><li><p>开发者创造合约环境 —— 平台运营</p></li><li><p>用户发送订单 —— 叫车</p></li><li><p>执行者执行订单请求 —— 司机接单</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/615148d4751ee9a704b6606e963d29b91a1979c603b25d0860ce5d6538da5ea2.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>用户端：Preference Environment（偏好环境）</p><ul><li><p>SUAVE本身就是多链架构，这给原本只能单链套利的小户提供了更多机会，因为肯定是多链跨链套利机会</p></li><li><p>本质解决了性能/工具的问题，单链低级别玩家只能在新手村活动，多链高级别玩家可以解锁更多游戏地图，SUAVE给了低级别玩家金手指去进入高端地图，消除了信息差</p></li><li><p>用户有特定的交易偏好，ask for 定制化服务，比如定制swap参数，指定从A链到B链进行交易，类似打车选择路线的时候指定司机按你规划的路线走高速之类</p></li></ul></li><li><p>执行端：Execution Market（执行市场）</p><ul><li><p>SUAVE引入了执行者（Executor）的角色，区块构建者、RPC服务提供商、钱包等等都可以当执行者</p></li><li><p>执行者之间良性竞争，根据用户偏好竞争接单上岗，就好像滴滴快的大战为了抢占市场打价格战一样，几个执行者也会为了抢单子把更多的MEV返还给用户（Minimizes MEV for users）</p></li><li><p>因为SUAVE是一个更公平的共享打车平台，mempool所有用户的偏好都是公开的（给了广大网约车司机参与接单的权利），所以解决了原本市场blockbuilder通过EOF独家订单流（地方出租车 like 纽约小黑车）信息差非对称博弈来获利的问题</p></li></ul></li><li><p>构建端：Decentralized Building（去中心化区块构建）</p><ul><li><p>上面两个环节闭环后（成功叫到车），由区块构建者的网络共同协作来构建完整的区块</p></li><li><p>构建者之间共享这些信息订单信息，在不泄露Orderflow和Bundle内容的前提下实现</p></li></ul></li></ul><p><strong>产品进度</strong></p><ul><li><p>前几周发了MEVM的文章<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://writings.flashbots.net/mevm-suave-centauri-and-beyond%EF%BC%8C%E8%AE%B2%E4%BA%86%E4%BA%A7%E5%93%81%E6%80%9D%E8%B7%AF%E8%A7%84%E5%88%92">https://writings.flashbots.net/mevm-suave-centauri-and-beyond，讲了产品思路规划</a></p><ul><li><p>修改EVM ，为 MEV 提供新的预编译（提供 MEV 供应链的每个底层作为预编译）</p></li><li><p>开发者可以在EVM中将 MEV 应用程序（如订单流拍卖或区块构建）编程为智能合约，因为很多其他东西SUAVE都帮你打包做了</p></li></ul></li><li><p>计划Q4推出 SUAVE Devnet，完成 Centauri 的发布，提供 MEVM 的初始版本</p><ul><li><p>执行节点：可信且私密的链下计算节点，可通过特殊预编译用于 SUAVE 上的智能合约</p></li><li><p>机密数据存储：存储机密数据供执行节点使用</p></li><li><p>Bids：包含用户想执行的机密数据以及允许访问机密数据的合约列表</p></li><li><p>SUAVE 链：用于智能合约部署和各方之间协调，未来还将托管用于支付偏好、预言机更新和数据可用性的资金。</p></li></ul></li><li><p>未来重点将放到用户的特定需求+跨域mev上</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/a7a4cb87fc5235a07389fe194a9623679dece18a36658e9d9a65804582a789ca.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">赛道图谱</h2><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/4d8b84f812c6cac7b5c0548aee706d85d0d105f95622b80984924cc29b22de92.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h2 id="h-pov" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">POV</h2><ul><li><p>未来值得研究的方向：</p></li><li><p>后ePBS时代协议外relay市场</p><ul><li><p>最近Ethereum Research文章指出协议外relay市场仍将继续发展</p></li><li><p>一方面，Relay可以提供更灵活的支付方式，因为其可以访问全部区块内容，并允许构建者以 ETH 以外的货币支付奖励。</p></li><li><p>不过由于 zkEVM可以允许builder在加密的有效载荷上发布竞价并准确支付提议者，relay的这一优势可能被削弱</p></li><li><p>另一方面，Relay也可以帮助降低连接的延迟，因为它可以充当proposer和builder的最快连接路径，通过TCP的方式而不是P2P网络</p></li><li><p>此外，ePBS 目前仍无法实现的竞价隐私，而Relay则可以选择不向其他builder透露竞价的价值以保护隐私</p></li><li><p>relay后续的几个潜在发展方向</p><ul><li><p>垂直整合的builder/relay：垂直整合 builder以降低relay提交区块的延迟和开销</p></li><li><p>relay即服务（RaaS): 考虑到relay已经拥有声誉、validator连接和运行infra的经验，其可以自己充当第三方实体向builder提供服务</p></li><li><p>Public goods relay：一些非审查（non-censoring）的中立性的relay，可以通过获取生态基金的公益性资助维持运营</p></li></ul></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/afe8ce8e569e0cbc33c457df2ebd0f1d86b44588bc7fe9e8ff551cb9b6f23986.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>PEPC</p><ul><li><p>PEPC，即 &quot;协议强制提案人承诺&quot;（Protocol-Enforced Proposer Commitments）最初是在 2022 年 10 月 Devcon 6 之前的 &quot;Unbundling PBS &quot;一文中提出</p></li><li><p>目的是作为一种庄严的协议小工具，允许验证者（&quot;提议者&quot;）对其生成的区块做出有约束力的承诺。</p></li><li><p>基于 EigenLayer 构建的框架，使validator能够做出区块排序承诺</p></li><li><p>遵循悲观执行方式，当一个区块违反承诺时不是slash惩罚，而是直接当成无效</p></li><li><p>通用化了 ePBS，允许提议者和一些构建者之间进行公平的交换，整体上还是偏理念落地有一段时间</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/cca850ac745bd409d86ec793b05eb448a5d2d318b86601c686d5088212aa0c53.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>Cross-domain MEV</p><ul><li><p>普通跨域 MEV</p><ul><li><p>通常在EVM L1和跨rollup之间发生</p></li><li><p>涉及到不同域的确认时间和最终行问题，如何协同和匹配</p></li><li><p>肯定是原本就跨域做block building的builder有话语权，可能有中心化趋势</p></li></ul></li><li><p>Shared Sequencer （SS）下跨域 MEV</p><ul><li><p>能够实现跨cross-rollup原子套利</p></li><li><p>要么两边执行交易发生了互换，要么就都不执行（能否实现有待验证）</p></li><li><p>感觉这个想象空间的前提是SS叙事真实且产生network effect</p></li></ul></li></ul></li></ul>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
        </item>
        <item>
            <title><![CDATA[friend.tech Research]]></title>
            <link>https://paragraph.com/@sallygu/friend-tech-research</link>
            <guid>u8ZO99KW47KL9GSBVBOS</guid>
            <pubDate>Wed, 11 Oct 2023 17:41:36 GMT</pubDate>
            <description><![CDATA[背景Friend.tech 前身是TweetDAO 和 Stealcam，设计人据称是@0xRacerAltTweetDAO2022 年 4 月建立，DAO创建者出售 Eggs NFT，允许持有者每天从 DAO 的 Twitter账户发一条推文售出的NFT基于旁氏模型，也是出售越多地板价越高目前已经被废弃，基本没什么人用，就是一个很原始的旁氏小盘子Stealcamarbitrum上的图片NFT分享社交协议让用户自己传图片，然后将其像素化生成NFT，其他用户要steal付费才能开图只有成为实际的owner才能开图，也是遵循旁氏模型，一个NFT越多人steal就越贵图片NFT每多被stolen一次，它价格就会自动上升0.001 ETH + 10%上次的价格本身图片的创建者和转手被stolen的人能分到钱，转手的可以拿回原款外+45%卖出利润，创建者和协议分别拿45%和10%相比起TweetDAO设计精致很多，更符合web2社交裂变的逻辑，缺陷可能在于本身图片形式的局限性在可玩性上，对比friend.tech差点意思，但协议细节基本是异曲同工简介基于Base生态的去中心化社交应用✖️社...]]></description>
            <content:encoded><![CDATA[<h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>背景</strong></h2><ul><li><p>Friend.tech 前身是TweetDAO 和 Stealcam，设计人据称是@0xRacerAlt</p></li><li><p><a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://tweetdao.com/">TweetDAO</a></p><ul><li><p>2022 年 4 月建立，DAO创建者出售 Eggs NFT，允许持有者每天从 DAO 的 Twitter账户发一条推文</p></li><li><p>售出的NFT基于旁氏模型，也是出售越多地板价越高</p></li><li><p>目前已经被废弃，基本没什么人用，就是一个很原始的旁氏小盘子</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/e3a324245c1fa841362e8cd544db36b3b4958146ae532b0efcbb30cb03dce0d0.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p><a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://www.stealcam.com/">Stealcam</a></p><ul><li><p>arbitrum上的图片NFT分享社交协议</p></li><li><p>让用户自己传图片，然后将其像素化生成NFT，其他用户要steal付费才能开图</p></li><li><p>只有成为实际的owner才能开图，也是遵循旁氏模型，一个NFT越多人steal就越贵</p></li><li><p>图片NFT每多被stolen一次，它价格就会自动上升0.001 ETH + 10%上次的价格</p></li><li><p>本身图片的创建者和转手被stolen的人能分到钱，转手的可以拿回原款外+45%卖出利润，创建者和协议分别拿45%和10%</p></li><li><p>相比起TweetDAO设计精致很多，更符合web2社交裂变的逻辑，缺陷可能在于本身图片形式的局限性</p></li><li><p>在可玩性上，对比friend.tech差点意思，但协议细节基本是异曲同工</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/7e51e0735e93525dff82c0b63a41f620e527e157bf8bc00020d170047af9b493.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/5f5810b58f97075f6302e4546c05d32141aef53bb76384622f94ddd1481553f0.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>简介</strong></h2><ul><li><p>基于Base生态的去中心化社交应用✖️</p></li><li><p>社交外壳下的金融交易轻应用 ✅</p></li></ul><h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>机制</strong></h2><ul><li><p><strong>注册方式</strong>：SMS、Google、Apple ID</p></li><li><p><strong>逻辑</strong>：有点像自己发币，然后通过FT找天使投资人</p></li><li><p><strong>需求</strong>：</p><ul><li><p>金融投机：前期投资/刮彩票赚潜在盈利机会，将平台视为赌场/交易所</p></li><li><p>名人社交：买入key后可以直接访问名人/KOL的内容、聊天室，给他留言，偶像经济那套</p></li><li><p>娱乐：蹭热度，多挂一层社交身份数据来源</p><p><code>由此可见，虽然是社交协议但其实没有很强的社交属性，是用伪社交需求包装的真金融产品</code></p></li></ul></li><li><p><strong>价格参数</strong>：</p><ul><li><p>采用统一的定价曲线：Price in ETH = supply^2 / 16000，即每股股份价格（ETH）=（股份总量^2）/ 16000</p></li><li><p>share数到达100，价格到达0.6ETH（1000U）</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/61c168fb390b905f1b99f3aef8deefb816edfff924f39c81d6a62d135d5a5268.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p><strong>底层资产</strong>：个人身份key/shares</p><ul><li><p>应该是基于对两个前身项目失败经验教训的总结，在底层资产的选择上做出了转型</p></li><li><p>不再用纯图片，而是用个人品牌和身份当作NFT的底层价值，从物到人拓宽了想象边界和平台上限</p><p><code>尽管依然秉持着对ponzi模型孜孜不倦的追求，但因为代表个人身份NFT的价值较为集中，所以可以预见整个旁氏走向终结的过程会在很大程度上有所延缓</code></p></li></ul></li><li><p><strong>使用流程</strong>：</p><ul><li><p>用户创建群/房间发行个人的权益NFT即key</p></li><li><p>发行人作为首位买家获得免费key</p></li><li><p>其他用户通过发现渠道购买key博取未来增长收益</p></li><li><p>随更多key持有者入场价格随曲线递增后，早期持有者卖出key套现</p></li></ul></li></ul><h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>特征</strong></h2><ul><li><p><strong>移动端应用</strong>：在手机打开官网并将其添加到主屏幕即可，无需appstore下载</p></li><li><p><strong>有进入门槛</strong>：需要获取邀请码才能进入dapp，并且进入应用前需要提前打至少0.01 ETH给Base 网络</p></li><li><p><strong>旁氏模型</strong>：key价格随数量增长，增速递增，越早进入回报系数越高</p></li><li><p>保留web2社交关联：比较取巧直接连推特，避免了社交网络和身份重构的麻烦</p></li><li><p><strong>思考</strong>：</p><ul><li><p>进入门槛的设置还是为了饥饿营销，比较像之前火过一阵的thread和clubhouse</p></li><li><p>整体做的就是一个很轻的产品，但只有移动端其实局限了</p></li><li><p>个人认为思路还是很奇怪，需要通过FT的链接点回X(twitter)，倾向于认为早期的流量的导入应该是从web2到web3而不是相反的路径，不过也可以是mutual的</p></li><li><p>连接推特但是没有关联好友，推特的原始内容数据也没有，所以保留社交资源保留了什么呢？</p></li><li><p>发现机制目前很糟糕透，没有推荐机制和良好的用户早期互动环节，如果放在真实的web2社交场景里去思考，产品思路似乎一开始瞄准的就不是陌生人社交而是熟人社交+bot社交？</p></li><li><p>根本上讲这类买key的投资行为并不是社交行为而是经济行为，基金给项目方投钱散户买股票等本质是交易</p></li><li><p>成功可能也是时代的产物，在无利不起早的年代里真实的利益出让才能打破社交壁垒</p></li></ul></li></ul><h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">改进提案</h2><ul><li><p>内置好友列表和关注账号上线通知</p></li><li><p>探索用户页面加入filter和推荐机制</p></li><li><p>为高阶玩家和大V提供自定义的定价曲线和社群配置</p></li><li><p>设置有效的bot监测和管理机制以防驱逐真实用户</p></li><li><p>拓展在web2更快捷有效的引流和推广渠道</p></li><li><p>划分不同的持有者等级并提供不同级别的定制化社交功能</p></li></ul><h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">商业模式</h2><ol><li><p>收10% 交易手续费，5%归协议国库 + 5%归key交易者</p></li><li><p>购买金额的90%存入底池提供流动性</p></li><li><p>积分空投和blur一样，预计未来 6 个月测试期间发1 亿积分</p></li></ol><h2 id="h-ui-and-ux" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">UI &amp; UX</h2><ul><li><p>检索页面</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/73e709174b84221a969b25817790018b2972fe5306ddc8ec86a5197a6e1761ed.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>购买key转账完成的页面</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/2d2b29f2056198cf333b2c237477609cbf38f898e772c66a525262825ee0212c.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>购买key后可以看到的留言互动页面</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/1fbb30727c8f61a2ba1dac680baa7263bfbc7b1dfcc1c5237c8ad89394899e37.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>网页端</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/0855ee5a86d6af497333e7687272c9399c3d2064524293e3825d47da768b0b31.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">融资</h2><p>Seed轮 - Paradigm - 估值未知</p><h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">运营</h2><ul><li><p>协议总费用达$20.6M</p></li><li><p>总交易量达$10.6M</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/b2146b9223512ed00c236952d1927586edd3dc0feba806dc98a7c86658895255.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>日交易量在逐渐回落，增长势头放缓</p></li><li><p>10月平均交易量在～130k</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/c9d59217b2296b060027ba738f62611c5f66bc540c8d91b8c5f9f6d1152e3354.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p>目前卖家数量超过买家，并且差距没有收窄的趋势</p></li><li><p>显示部分运营风险</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/45fe3b6f8811598f58a8b8240da5f87ce0ba91d76386445fb65eaa2b6854f9cd.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p><a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://dune.com/cryptokoryo/friendtech">https://dune.com/cryptokoryo/friendtech</a></p><h2 id="h-" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">竞品/仿盘</h2><ul><li><p><strong>Avalanche - Stars Arena</strong></p><ul><li><p><a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://www.starsarena.com/?ref=bityingmu">https://www.starsarena.com/?ref=bityingmu</a></p></li><li><p>官方孵化 关注度很高 29k followers</p></li><li><p>网页端，整个twitter集成和登陆过程更简洁快速</p></li><li><p>强调金融属性，对社交功能做了最大程度简化，但使用体验实际有所提升</p></li><li><p>生态位好，但有过黑客攻击历史，运营风险需要关注</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/a7685889be3507b130c8de9a4730ed1c0552cc3763039e9e016ac8163a1e9c2f.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/70837dddab113463a56047a7e5e6dd6f7f3268a697ab6b53fbc0ca0a1b55124b.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p><strong>Arbitrum - Cipher</strong></p><ul><li><p><a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://cipher.rip?ref=bityingmu">https://cipher.rip?ref=bityingmu</a></p></li><li><p>网页端，UI UX做了美化提升</p></li><li><p>更像twitter有公共论坛但有资金门槛5core</p></li><li><p>发现机制做了改进单其他没有太多创新</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/b5829748401076cac413f922fab338741c9244986c5198bc515260a62209240f.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/552729da019e9e0e68374ce8985fb6e274a65a72bda38fee6aaeb5feddb2bd7f.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p><strong>BTC L2 - New Bitcoin City</strong></p><ul><li><p><a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://newbitcoincity.com/app">https://newbitcoincity.com/app</a></p></li><li><p>移动端，需要设置私钥，BTC本位，17k followers</p></li><li><p>也是强化金融交易属性，有(3,3)锁仓30天的机制</p></li><li><p>value-add的功能比较多比如语音、图片发布、专业交易工具</p></li><li><p>其实对BTC生态来说这类产品很合适去打web2用户市场，有想象空间</p></li></ul></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/116065814d17209cc23885a6739fe05c9f8eb4c1bcf1bccaa4d80966b5ee82f9.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ul><li><p><strong>Mantle - Fan.tech</strong></p><ul><li><p><a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://fan.tech/">https://fan.tech/</a></p></li><li><p>移动端，更偏社交属性，还未上线</p></li></ul></li></ul>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
        </item>
        <item>
            <title><![CDATA[长风破浪会有时：迈步NFT 2.0]]></title>
            <link>https://paragraph.com/@sallygu/nft-2-0</link>
            <guid>IE6bhQ1xKc3yVeQidu99</guid>
            <pubDate>Fri, 09 Sep 2022 08:05:13 GMT</pubDate>
            <description><![CDATA[中本聪认为比特币是一个自我实现(self-enforcement)的预言。以此类推，如果NFT能够和比特币一样实现信念和选择的同调，我们也有理由认为NFT的价值可以在远期的纳什均衡中被持续加固。当然信念和选择的锚定是浮动的，目前NFT远远没有达到比特币的共识程度，自然很难去控制或确保这种锚定。 想要解决这个问题，意味着我们仍然需要增加支撑纳什均衡等式的条件——NFT的内在价值和使用价值。道理其实非常容易理解，郭敬明老师就曾在他的青春疼痛文学作品中告诉我们：”没有物质的爱情就像一盘散沙”，高度提炼了唯物主义价值观的精髓。所以兴趣和信念仅仅是必要不充分条件，只有实实在在的价值提升了，需求才会刚性，选择才会坚定。 不难看到，目前市场上已经演化出两种打开需求侧的思路：向右迈步提升投资属性——NFT金融化（Financialization）向左迈步提升消费属性——NFT商品化（Commoditization）接下来本文会对这两种思路分别展开进行论述。1. NFT金融化：DeFi x NFT = NFT FiNFT金融化并不是一个新奇的话题，早在2021年上半年我们就已经看到许多围绕这个叙事...]]></description>
            <content:encoded><![CDATA[<p>中本聪认为比特币是一个自我实现(self-enforcement)的预言。以此类推，如果NFT能够和比特币一样实现信念和选择的同调，我们也有理由认为<strong>NFT的价值可以在远期的纳什均衡中被持续加固</strong>。当然信念和选择的锚定是浮动的，目前NFT远远没有达到比特币的共识程度，自然很难去控制或确保这种锚定。</p><p>想要解决这个问题，意味着我们仍然需要增加支撑纳什均衡等式的条件——NFT的内在价值和使用价值。道理其实非常容易理解，郭敬明老师就曾在他的青春疼痛文学作品中告诉我们：”没有物质的爱情就像一盘散沙”，高度提炼了唯物主义价值观的精髓。所以兴趣和信念仅仅是必要不充分条件，<strong>只有实实在在的价值提升了，需求才会刚性，选择才会坚定。</strong></p><p>不难看到，目前市场上已经演化出两种打开需求侧的思路：</p><ol><li><p><strong>向右迈步提升投资属性——NFT金融化（Financialization）</strong></p></li><li><p><strong>向左迈步提升消费属性——NFT商品化（Commoditization）</strong></p></li></ol><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/97e7f377eb08a074997009f191641c998bd7e857f0813881638858ad2d3536bb.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>接下来本文会对这两种思路分别展开进行论述。</p><h2 id="h-1-nftdefi-x-nft-nft-fi" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">1. NFT金融化：DeFi x NFT = NFT Fi</h2><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/472bebf2f7ef7087fe0d10f90b89f89cd190001712cdbf13592e8f9aa4c1ff8a.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>NFT金融化并不是一个新奇的话题，早在2021年上半年我们就已经看到许多围绕这个叙事的讨论（详见<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://mirror.xyz/0x5Eba828AB4999825D8416D7EAd9563b64FD90276/S7S3rAqnUyfJ2g9M17yeNruqAfgftfQRDH1zLAgMt_c">IOSG</a>, <a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://medium.com/1kxnetwork/fungible-non-fungibles-the-financialization-of-nfts-32565adf454a">1kx</a>, <a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://www.hashkey.com/viewerjs-0.5.8/documents/cn/blockchain_research/%E3%80%90%E5%8C%BA%E5%9D%97%E9%93%BE%E7%A0%94%E7%A9%B6%E7%AC%AC158%E6%9C%9F%E3%80%91-NFT%E9%87%91%E8%9E%8D%E5%8C%96%E6%A1%86%E6%9E%B6.pdf">Hashkey</a>等研报）。因为本身NFT就是一种链上的代币标准，包含ERC20，ERC721，ERC1155等几种协议，所以将其视为一种投资品，并想进一步延展探索其金融属性和衍生品是非常直接和鲁棒的逻辑链传导。那么问题就在于:(1)NFT金融化能不能解决需求？(2)NFT金融化到底能多大程度上打开NFT需求侧？</p><p>我们认为NFT的金融化可以很大程度上帮助拓展和提升需求，主要基于以下两点:</p><ol><li><p><strong>通过结合DeFi玩法提升NFT共识</strong>。NFT在与DeFi协议结合后，可以降低参与NFT市场的门槛和教育成本。因为对非传统金融工具的交易进入门槛相对较高，玩家在缺乏对这种独特或奇异资产的知识信息时，往往会承担更高的交易风险，从而阻碍了更多长尾买家进入市场。而缺乏足够的交易深度和市场信心，自然会延缓NFT整体价值共识的累积。</p></li><li><p><strong>通过证券化提升可交易性并激发流动性</strong>。通过将非流动性的NFT资产转换为可交易证券，NFT可以像传统证券资产一样被用作抵押品等帮助切割和分散基础资产池和信用风险。而NFT资产及其衍生交易品的诞生与应用，Web3参与者就可以通过分仓的方式将风险分散到更广泛的类目中，从而使自己获益。在以上认知的基础上，我们认为可以首先重点关注NFT Fi赛道中如下3个方向：</p></li></ol><p>在以上认知的基础上，我们认为可以首先重点关注NFT Fi赛道中如下3个方向：</p><ul><li><p><strong>Exchanges：建立NFT即时交易对</strong></p></li><li><p><strong>Collaterals：将NFT用作抵押品</strong></p></li><li><p><strong>Derivatives:：发展NFT衍生品</strong></p></li></ul><h3 id="h-a-exchanges" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">A. Exchanges</h3><p>类似Opensea、LooksRare样的传统挂单交易所存在的一个很大瓶颈就在于其无法为NFT提供即时流动性，进而无法进一步提升资本效率。而解决这一问题最直白且有效的思路就是搭建NFT（ERC-721）和FT（ ETH/ERC-20）的瞬时交易对。</p><p>AMM毫无疑问是实现这一思路的极佳形式。过去典型地如<strong>NFTX、NFT20这类碎片化的NFT流动池解决方案</strong>，已经成功实现了同系列NFT藏品的价值汇聚与平均化处理。但问题在于其单个NFT由固定数量的FT表示，<strong>近似于uniswap V2的x*y=k 恒定乘积(constant product)算法</strong>，而NFT与FT在流通量上相差悬殊，因此很容易会导致NFT市场深度不足和高滑点。</p><p>以<strong>Sudoswap为代表近期新兴出现的另一种升级版AMM机制</strong>在很大程度上改进了这一问题。一方面，其允许LP通过恒定、线性或指数形式的<strong>联合曲线（Bonding Curve）自行设置参数变量delta</strong>，从而在报价模式上解放了自由度，控制住了大额订单滑点的恶化。另一方面，与NFTX等传统AMM不同的地方还在于Sudo省去了将NFT兑换为ERC-20的过程，并且玩家可以自行选择设置单边或双边交易池。因而这类基于联合曲线的AMM之于恒定乘积的AMM可以被认为是Uniswap的V3之于V2。</p><p>不过目前市面上的NFT AMM机制仍然只是简单地将NFT同质化为FT，<strong>很大程度上只适用于中长尾的游戏类NFT资产</strong>。对于属性特征明显藏品价差较大的头部NFT来讲远远算不上令人满意的流动性解决方案。同时<strong>在用户操作流程和联合曲线的设置上，现有平台的处理都十分粗糙</strong>，期待其未来可以添加一些简单的阶跃函数和用以参考的数据分析界面。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/53902b307660c8bf673a8b69cd2a8807c0be16ce16e1453fcf9f1addce17b258.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h3 id="h-b-collaterals" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">B. Collaterals</h3><p>考虑到目前主流的NFT借贷实际都是基于蓝筹的pfp NFT，我们可以先以传统的艺术藏品借贷市场用作参考。历史上看，通过将艺术藏品抵押获得贷款的金融范式在传统银行业中存在已久。<strong>而艺术藏品贷款产业的估值早已占到全球艺术品市场总价值的</strong><a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://www.ubs.com/global/sc/media/display-page-ndp/sc-2"><strong>四分之一</strong></a>。德勤和ArtTactic在去年的<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://www2.deloitte.com/content/dam/Deloitte/lu/Documents/financial-services/artandfinance/lu-art-finance-report-2021.pdf">Art&amp;Finance报告</a>(p193-202)中预估全球艺术藏品抵押贷款总额已超240亿美元，同比增长10.7%。此外，过去的<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3099054">相关研究</a>也指出<strong>在经济萧条时期，艺术藏品的借贷需求反而会基于地缘性出现上升</strong>。</p><p>那么由此引发的下一个问题就是，<strong>这种抵押借贷的需求由何而来？</strong></p><p>我们认为相对代币而言，市场进入者对于直接出售手中蓝筹NFT藏品的意愿往往更低。然而长持NFT也意味着大量资金会被锁在非流动性资产中。因此一方面，在市场环境较好时，持有人容易发现更好的短期投机机会，自然地，将NFT用作抵押品临时出借释放流动性会是一个优良选项；另一方面，当经济环境较差时，持有人会倾向于利用类似Compound, Aave等DeFi协议中存在的超额抵押方式，在确保不失去手中NFT的同时持有更多现金来预防潜在或突发的短缺危机（这也解释了为何总体NFT借贷的违约率较低）。</p><p>一般地，我们可以将目前市面上的NFT抵押借贷解决方划分成<strong>点对点（P2P）, 点对池（P2Pool）,抵押债仓（CDP）和场外借贷（OTC）四类：</strong></p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/ef732d238c1268e68d700c53790b499018da0c3c5cd152e0ba5cbe66aa13a4ad.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h3 id="h-c-derivatives" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">C. Derivatives</h3><p>金融衍生品即<strong>围绕标的资产（Underlying Asset）所产生的一种合约或金融工具</strong>。常见的金融衍生品包括远期（prediction）、期货（futures）、期权（options）等等。通常金融衍生品都具备高风险高收益的特性，并被广泛用于对冲和分散风险。以房产为例，买家可以在不持有原生资产（实际房产）的情况下买入顶层（senior tranche）CDO看多，但同时也面临违约可能性。针对NFT这类标的资产，我们也观察到目前市场上正不断涌现新兴的衍生品协议，试图进一步激活NFT市场流动性并降低交易成本。简单地，我们可以将NFT金融衍生品分为<strong>期货、期权、保险和结构性产品</strong>这四类。</p><ul><li><p>以NFTprep, NFTures为代表的NFT期货方案可以帮助实现针对NFT资产的长短交易，并推高盈亏杠杆。但这也对预言机的价格喂送和做市商的供池深度提出了更高的要求。目前市面上主要产品都还在测试网的阶段。</p></li><li><p>以JpeX，Nifty等为代表的NFT期权方案允许用户通过购买看涨（call）或看跌（put）期权来对冲手中NFT的地板价波动，并在一定程度上降低了NFT的投机门槛。但目前几个项目也都处于早期，面临流动性困境。</p></li><li><p>为稀缺/高净值或代表链下资产的NFT进行投保，也是一种值得探索的风险转移方式。考虑到NFT的特性需要在传统保险模式的基础上进行一定调整。</p></li><li><p>类似Index Coop JPG这样根据不同NFT或NFT产品建立起来的指数基金，以及Cyan, Cedar等为代表的先买后付项目，都是对NFT结构化产品不错的探索。</p></li></ul><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/ba72eb8e3c33e3af2dcdb981c2dade1cc78247192ad51bb3a42cd7ee25f14a87.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>在<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://mp.weixin.qq.com/s/lO7W3xOhDxctvCgG6i_9qg">过去的研究</a>中，我们已经对上述提及的部分借贷和衍生品项目做过解读和梳理，更多内容将会在未来的《NFT Fi Report》进一步展开论述。</p><h2 id="h-2-nft-x-nft-nfg" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">2. NFT商品化：新消费 x NFT = NFG</h2><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/2b84c10c9e162e7cbbd78ce7780d12a74db4e00dd2d3712a6ca7b5aea85c95b1.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>NFT商品化是一个很有意思的话题，因为这也许意味着对整个大消费赛道的颠覆。但在讨论NFT商品化之前，我们有必要首先回答的一个问题是：</p><p><strong>NFT究竟能不能被视为传统商品？</strong></p><p>从根本上讲，消费品的存在是为了解决和镇静马斯洛金字塔中广泛存在的诉求和烦恼。**人们购买和消费大宗商品的首要目的必定不是为了投机或获利，而是为了利用其特有的功能性价值，来缓解自己短期内的某种痛苦。**举几个简单的例子，我们买米是为了果腹，买车是为了代步，买游戏是为了娱乐，买咖啡是为了提神。由此逆推，当一件商业化的物品被人购买的主要原因是想利用和交换其本身的功能属性，而非谋求盈利回报时，它理应可以被视为一种消费品。</p><p>遵照民众消费的底层逻辑进行逆推，并结合当下的时代背景，我们可以做出一个待验证的假设，即如果NFT能够实现：</p><ol><li><p><strong>首要购买目去投机化</strong></p></li><li><p><strong>内生文化的大众认同</strong></p></li><li><p><strong>功能品类(assortments)的规模覆盖</strong></p></li></ol><p>就可以被嵌套进传统商品和消费品的叙事逻辑中。而如果可以把NFT直接和传统商品划上约等于号的话，一个值得探讨、但不一定对的想法将会是：把Web2电商的投资逻辑平移至web3的NFT marketplace赛道。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/91c50dcee450170bee252658a78dbf5d13820b780e5add7a1f96f584cda97b61.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>那么回归正题，究竟NFT商品化的发展有哪些方向？能代表哪些使用场景？ 我们认为凡是具备 <strong>(1)非完全同质（2）受益于所有权数字凭证</strong> 这两类属性的任何产品其实都可以依托于NFT的形式进行表现。并且由此引发的一个有趣的概念将会是<strong>非同质化商品（Non-Fungible Goods, NFG)</strong>。这种虚拟资产与实体经济相结合的叙事概念，可以在很大程度上进一步拓宽NFT的实际使用价值和未来应用场景。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/f89eb00763571c9dedc21f55c6fe3a0529a62fd4119214deb211293790fa830f.png" alt="NFT相关商品化用例（ @shivsakhuja ）" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">NFT相关商品化用例（ @shivsakhuja ）</figcaption></figure><p>简单地，我们可以根据NFG针对所具备的不同将其划分为<strong>私有型（Private Goods）和</strong>公共型（Public Goods）<strong>两类。其中私有型又可以细分为</strong>可转让(Transferable)与不可转让/灵魂绑定（SBT）<strong>两种范式，公有型主要以</strong>CC0/<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://a16zcrypto.com/introducing-nft-licenses/">CBE</a>(can’t be evil)叙事为代表进行讨论。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/051bd6deafbef852fd4b5bcf45895425fb12d490ce5e76e4f4ef66036070dbde.png" alt="NFG Category" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">NFG Category</figcaption></figure><h3 id="h-a-private-goods" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">A. Private Goods</h3><ul><li><p><strong>可转让（Transferable）</strong></p><p>典型可转让的NFG用例即门票和通行证。 由于通常演唱会、体育比赛及其他大型活动门票的数量都有限制，并且访问场馆需要出示相关凭证才能获得进入许可。因此以NFT为依托是一个非常合适的选择。 <strong>一方面NFT门票可以在最大程度上消除伪造问题</strong>，使得验证门票的所有权和真实性变得更为容易；<strong>另一方面，NFT化门票也可以极大激活二级市场的交易流动性</strong>。在Ticketmaster或Vividseats等类似传统web2门票二手平台上，市场交易的效率较为低下，并且买方的权益很难得到保障。NFT化门票则可以帮助建立起一种公平的保护双方权益的标准化市场。 此外，通过收集门票持有人在链上的历史归属数据，表演者或主办放也可以在未来考虑空投通行证、折扣、赠品等给粉丝进行回馈。</p></li><li><p><strong>灵魂绑定（SBT）</strong></p><p>不可转让的NFT（SBT）可以被广泛地用于充当去中心化数字身份证明（DID）或可验证凭证（Verifiable Credential）。考虑到本质上NFT就是防伪且任何人都可以验证的数字证书，那么将学历证书、技能证书、文凭、分数和成就等通过SBT NFT的形式实现发布和持有，就意味着可以让任何人核实其合法性，从而实现用户身份的去信任化验证。而其中类似Sismos这样基于zk证明技术为用户发放SBT（ERC-1155）身份徽章的协议，还能够在保护用户数据隐私的基础上完成闭环。 更进一步地，SBT NFT还可以被用来解决验证选票这样的问题，并有可能成为全球性的身份标准，帮助消除区域间因身份认证系统互不兼容而产生的摩擦与隔阂。</p></li></ul><h3 id="h-b-public-goods" class="text-2xl font-header !mt-6 !mb-4 first:!mt-0 first:!mb-0">B. Public Goods</h3><ul><li><p><strong>CC0/CME</strong></p><p>日本社会学家三浦展在《第四消费时代》中将日本消费社会的发展划为四个阶段，指出民众的需求也和西方一样正在经历由基础性消费，到追求个性品牌与文化消费的演化。而这一现象正在逐步延展到全球许多发展中国家。越来越多的人正在跨过基础消费层，对商品的品牌文化的重视取代了对性价比的追求。</p><p>而 CC0（Creative Commons Zero）这种代表作者放弃一切IP版权的协议形式，毫无疑问能够进一步激发人们对于这类项目品牌文化的追捧以及二次创作热情。与传统私有化品牌创作的NFT商业模式相比，CC0通过将商品带入公有领域而使其逐渐演化为了一种<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://a16zcrypto.com/cc0-nft-creative-commons-zero-license-rights/%E3%80%81">开源并具备网络效应的“应用程序”或’“平台”</a>。在衍生品被不断创作和分享的正循环下，原作品牌所获得注意力以及社区内部的集体共识自然可以获得延长和加固。这也不难解释为何在运营难度大大增加的情况下，仍然不断有NFT项目宣布转向CC0模式。</p></li></ul><h2 id="h-closing-thoughts" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">写在最后Closing thoughts</h2><p>NFT是一个浪漫又理想主义的概念，罗曼罗兰中又夹杂了一些聂鲁达。但诗意并不能解决我们现实中的问题，也无法提供形成强共识的基础。NFT的发展需要更多的马基雅维利去实实在在地提升用户的边际效用，否则NFT2.0的前进也许只能止步于空想。</p><p>从亚当斯密的生产消费论，到阿尔弗雷德马歇尔和哈特的“消费者主权”(Consumer’s Sovereignty)说，再到哈耶克“消费者对生产者主权行使”的诠释，我们看到随着互联网技术的发展，消费者对个人自身偏好完全实现越来越强烈和明确的诉求。</p><p>18年前，克里斯安德森（Chris Anderson）在《连线》发表了经典的“长尾理论”(<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://www.wired.com/2004/10/tail/">The long tail</a>)，其中就指出<strong>关注成本在网络时代的降低，使得在正态分布曲线尾部的个性化和分散的长尾需求，有可能累加起来形成一个比流行需求更大的市场，并且经济效益甚至会超过“头部”（the hits）</strong>。这意味着面向小众细分市场的利基商品很可能蕴藏着更大的价值和增长潜力。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/81e55cd204991dada8bd9b0dc9220641ce760aed42165f0458f8644bc5a5b3fc.png" alt="The Long-tail Theory" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">The Long-tail Theory</figcaption></figure><p>相信在金融化和商品化两步齐迈的背景下，NFT的需求侧增长会实现进一步的飞跃。而<strong>NFT这类利基商品有机会在未来爆发出更大的价值增量，在真正意义上改变web3的进程和人们的生活方式。</strong></p>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
            <enclosure url="https://storage.googleapis.com/papyrus_images/13b33f0e0a285c63459480b1ec9667655cb34b442c14eaa666654537f102d5b5.jpg" length="0" type="image/jpg"/>
        </item>
        <item>
            <title><![CDATA[NFT炼金术：定价Pricing 101]]></title>
            <link>https://paragraph.com/@sallygu/nft-pricing-101</link>
            <guid>xHZeo3LVlDvdTHNBkJYA</guid>
            <pubDate>Tue, 26 Jul 2022 07:00:17 GMT</pubDate>
            <description><![CDATA[NFT如何定价一直是一个有趣的话题。因为定价是一个不可避免的中间操作，包含着可计算和不可计算的两部分问题，在任何NFT Fi的应用场景下都需要解决。想要将NFT普遍应用到DeFi中，彻底激活NFT的流动性，首先我们需要对NFT这一资产价值做出尽可能真实且能被广泛接受的估值判断。 然而NFT由于以下三个方面，难以像传统金融资产一样被简单进行评估计算： 1.NFT非同质化属性所导致其内生的主观性和非流动性 2.NFT的稀有度相对模煳，并且稀有度和价格水平不完全正相关 3.NFT价格波动剧烈（受团队和政策问题影响经常发生暴拉和暴跌） 而如果定价机制无法得到很好的解决，那么NFT贷款交易等行为往往会因高风险而难以赢得市场信任，进而导致两个问题： (1 )缺乏支撑交易池深度的充足流动性; (2)难以构建NFT形式的多样化金融衍生品。 为了解决这一问题，市场上越来越多NFT定价平台和新兴方法正在涌现。在这里我们可以简单将这些定价解决方案分为两类: 同行定价类: 又可细分为 (a)主体评估定价和(b)流动池博弈定价预言机定价类：又可细分为 (a)TWAP定价和(b)链下计算定价1.Peer同行...]]></description>
            <content:encoded><![CDATA[<p>NFT如何定价一直是一个有趣的话题。因为定价是一个不可避免的中间操作，包含着可计算和不可计算的两部分问题，在任何NFT Fi的应用场景下都需要解决。想要将NFT普遍应用到DeFi中，彻底激活NFT的流动性，首先我们需要对NFT这一资产价值做出尽可能真实且能被广泛接受的估值判断。</p><p>然而NFT由于以下三个方面，难以像传统金融资产一样被简单进行评估计算：</p><p>1.NFT非同质化属性所导致其内生的<strong>主观性和非流动性</strong> </p><p>2.NFT的<strong>稀有度相对模煳</strong>，并且稀有度和价格水平不完全正相关</p><p>3.NFT<strong>价格波动剧烈</strong>（受团队和政策问题影响经常发生暴拉和暴跌）</p><p>而如果定价机制无法得到很好的解决，那么NFT贷款交易等行为往往会因高风险而难以赢得市场信任，进而导致两个问题：</p><p><strong>(1 )缺乏支撑交易池深度的充足流动性;</strong></p><p><strong>(2)难以构建NFT形式的多样化金融衍生品。</strong></p><p>为了解决这一问题，市场上越来越多NFT定价平台和新兴方法正在涌现。在这里我们可以简单将这些定价解决方案分为两类: </p><ol><li><p><strong>同行定价类: 又可细分为 (a)主体评估定价和(b)流动池博弈定价</strong></p></li><li><p><strong>预言机定价类：又可细分为 (a)TWAP定价和(b)链下计算定价</strong></p></li></ol><h1 id="h-1peer" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">1.Peer同行定价</h1><h2 id="h-acrowds" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">a.Crowds主体评估定价</h2><p>主体评估定价是目前主观性较强的一种定价形式。以Taker V1为代表的流动性借贷协议，可以通过将DAO的利益与贷款人的利益相结合，以主体评估和投票决策的形式对NFT进行白名单划分和定价，从而在一定程度上对贷款人所面临的风险敞口进行缩减。同时，这种方式对NFT资产的质量要求没有限制，可以被广泛应用于长尾和新兴NFT藏品的价格发现。然而这种方式对curator的判断能力依赖严重，并且其无法提供实时更新的NFT价格，整体效率较低。</p><h2 id="h-taker-v1" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Taker V1</h2><p>Taker是一个NFT资产的流动性协议，主要通过DAO的形式来为NFT借贷提供流动性，支持包括NFT，证券，合成资产等在内的多种形式的资产。持有TKR代币后可获得DAO成员资格，参与决策借贷利率和公平定价等设计。同时，持有TKR也能通过质押获得额外收益。</p><p>Taker DAO内部设有多个curator DAO(sub-DAO), 每个sub-DAO可以自行管控自己的白名单及白名单上任何NFT的地板价以防借款人违约。此外，sub-DAO的成员通过集体投票，决定将自有国库的资金投入特定类型的NFT资产上。比如有些sub-DAO可以只关注Metaverse土地类资产，有些sub-DAO可以只关注pfp艺术类资产。</p><p><strong>在Taker群体定价的模式下，完整的借贷流程如下：</strong></p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/81c1661d9773aec3ba5488f26ba61eb28d7ce7d15ef0b6af64c977151a31dbe5.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ol><li><p>Taker社区成员（贷款人）向DAO存入资金。</p></li><li><p>DAO铸造DAO代币（TKR）以代表成员的股份</p></li><li><p>Curator（发起人）对所发起的DAO中NFT藏品进行主观定价</p></li><li><p>借款人以NFT为抵押，根据Curator的评估价格进行贷款</p></li><li><p>借款人用利息偿还贷款</p></li><li><p>DAO成员获取奖励（根据利息收益率）</p></li><li><p>随着DAO的发展壮大，DAO成员获chi得的收益累增** **</p></li></ol><h2 id="h-b-spot" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">b. Spot流动池博弈</h2><p>流动池博弈定价机制和defi类似，主要是通过乐观的质押凭证机制，即由流动性提供者根据自己对价格的预期进行质押，从而将NFT的估值和流动池内的资产价格进行绑定。这也是一种很大程度上依赖于LP主观决策的定价方式，优势是其可以实现NFT的实时估值，将交易价值与真实价值挂钩，释放出更大的流动性。但其也存在定价机制复杂，不适用于对低价值的长尾NFT进行大量快速定价等问题。</p><h2 id="h-abacus" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Abacus</h2><p>Abacus是一个简单清晰的NFT估值系统，主要利用乐观PoS来创建一个以流动池为基础的NFT估值方法。Abacus的估值方法分为两种，一种是我们上述提到的群体定价，另一种就是我们这里所讨论的流动性定价。其流动池内的价值折算成ETH就相当于是开池NFT的价值。在这种机制下，相当于制定了一套ETH/NFT的交易对，可以像Opensea一样对NFT价格进行实时反映。</p><p><strong>基于其流动池定价方法，完整的借贷流程如下所示</strong>：</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/9d8a74364cf3c4339e51a3bf13d24705d8391f9b7ad7a0d2a3d9c32cd55d39f8.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ol><li><p>开启池子：</p><ol><li><p>NFT非托管，所有者需要在池子上签名作为凭证(proof of life)</p></li><li><p>所有者将收到一个NFT(ERC721)代币来代表他们的财产并赚取交易费</p></li></ol></li><li><p>交易者将ETH锁在池中</p><ol><li><p>决定锁定的ETH的数量：如果资金池是2eth，但交易者认为NFT价值2.5eth，交易者可以把0.5eth放入资金池，现在资金池价值2.5eth。</p></li><li><p>决定锁定时间：如果交易者认为价格只会在当前价进行短暂停留，可以只锁定2周。相反，如果交易者对底价有信心，可以锁定更长的时间，从而获得更多的奖励。</p></li></ol></li><li><p>NFT所有者开启释放</p></li><li><p>NFT所有者发送贷款请求</p></li><li><p>所有权被转让给借贷平台</p></li><li><p>借贷平台检查现货规模和锁定时间</p></li><li><p>借贷平台发放贷款</p></li><li><p>一旦借款人未执行还款，NFT将被立即拍卖</p></li></ol><h2 id="h-2-oracle" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">2. Oracle预言机定价</h2><h2 id="h-a-twap" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">a. TWAP</h2><p>预言机对NFT进行定价评估的典型方式就是基于简单的传统算法交易策略，在对NFT销售价格和底价进行加权平均计算，从而综合得到一个TWAP(Time Weighted Average Price)。举一个简单的例子，如果我们想计算时间间隔为1个小时的TWAP价格，就可以取开始和结束的累计价格P1和P2的差值，以及开始时间T1和结束时间T2的差值，将两者相除，从而算出这1小时内的TWAP。最为人所熟知的TWAP预言机包括Chainlink和Uniswap的V2。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/69f8cf096d7337d2974f96c0ea50a75e6254fb5ce89c41a60940ee890e860c96.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>实际上TWAP定价是最容易实现和最高效的一种方式，只要通过集成、爬取和清洗NFT交易平台的价格数据，在设定的时间序列内选取多个价格取均值就可以降低恶意操纵的可能，提供一个能被接受的、相对准确的NFT价格。</p><p>然而TWAP并不是一个完美的解决方案，因为在极端市场环境下，当价格出现剧烈波动时，TWAP预言机就很容易受到影响而变得不准确。因此TWAP被认为只适用于对市场活跃度较高，流动性较好，价格相对稳定的蓝筹NFT进行定价。</p><h2 id="h-benddao" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">BendDAO</h2><p>BendDAO是解决NFT流动性问题的借贷协议，借款人可以通过抵押NFT借出ETH套现。目前BendDAO可以支持包括BAYC、Cryptopunks、Azuki、MAYC、CloneX、World Of Women、Coolcats、CyberKongz和Doodles在内的9种蓝筹NFT的借贷。</p><p>BendDAO采取的NFT定价方式就是典型的TWAP定价。其通过和chainlink合作，调用多节点采集所抵押的NFT在Opensea和LooksRare两个交易平台的地板价，调用合约接口将地板价喂到链上后计算出对应的TWAP，从而过滤掉交易平台价格波动的影响。由下图可见对cryptopunk这一藏品，预言机提供的地板价和成交均价、TWAP是保持一致的。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/ab9284e039bf17ab48dc5dc7f5eeb9c7b6839549b58c2e070433a2b3e55c4e86.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>与BendDAO类似，运用TWAP预言机进行定价的协议还包括JPEG’d, DeFrag, DropsDAO，Pine等。</p><h2 id="h-b-off-chain-computation" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">b. Off-chain computation链下计算</h2><p>基于AI和机器学习的链下计算也逐渐成为一种新兴的NFT预言机定价方式。由于NFT的非同质性，其主要的属性分类、稀有特征、历史销售数据等价值信息可以通过元数据分解用作模型指标。协议进而可以基于这一系列指标和数据集进行建模处理，从而给出一个相对可靠和准确的定价或定价区间。</p><p>这类估值方式具备极高的技术壁垒，对于长尾NFT资产相对友好，可以被认为是最具备大规模应用可能性的解决方案。但问题在于这种方法对算力和元数据要求较高，由于算法未公开我们也无法确定训练和拟合结果是否有效。并且NFT一旦属性特征发生变化，模型很可能就会失效，因此需要不断迭代。</p><p><strong>Banksea</strong></p><p>以Banksea为代表的预言机协议主要采用 AI 模型来训练 NFT 的数据集，从而针对不同NFT资产生成准确、高效的预测价格。其整体主要由采b层和NFT层的两个模组构成。</p><p>在采集层上，Banksea会收集链上的NFT交易和上架记录实时计算出三类价格：市场地板价、AI地板价和24小时均价。AI地板价代表的是所有AI估值中的最低价，在市场出现剧烈波动或遭遇预言机攻击时起到一个风控和维稳的作用。</p><p>在NFT层上，Banksea会通过提取NFT的多维度特征，基于时间序列进行AI模型训练，定期生成标准估值和估值范围两个结果。此外，其会将AI模型计算得出的估值，与实时交易价格进行拟合和回归验证，从而优化最终结果并缩小误差区间。</p><p><strong>Banksea基于AI模型的链下定价流程可见如下</strong>：</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/ac01d30acefacfbe0806f989408bb60c80eaf4912688c638d4f59bce9890bf6e.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><ol><li><p>外部API查询：监控和抓取综合NFT数据，来源包括交易平台、社交平台、公链和抵押平台等</p></li><li><p>数据聚合：将收集得到的NFT数据进行清洗，提取特征属性，输入到AI节点</p></li><li><p>AI建模：AI节点集群根据输入的数据集进行模型训练和部署，计算出预测价格和风险分数，并将结果返回Banksea智能合约</p></li><li><p>数据提交：链上智能合约去除异常值后，提取出在合理区间内的数据提交给第三方程序</p></li></ol><p>除了Banksea以外，Upshot和NFTBank等也提供基于AI中更细分的机器学习(ML)方法对NFT进行精准定价的预言机解决方案。此外如Defi Kingdom、Axie Infinity等都集成AI链下算法进行定价的社区工具。</p><h1 id="h-one-more-thing" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">One More Thing</h1><p>最后，我们可以将目前市面上NFT定价两大范式中的四种具体解决策略通过下述几个维度进行概括梳理：</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/1f00307c339b492997fb283c172dc5fb7b376e95e4970c6ed39d364ed5e7fcab.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>可以看到，目前不管是哪种估值方式，都存在一定的优缺点。我们期待着在不远的将来更多新兴的NFT定价方式被挖掘和完善。尤其是对于链下计算的预言机定价方式，我们相信随着技术进步和更多优质项目方的参与，<strong>更多的AI算法技术如深度神经网络（DNN）等可以被投入到拟合评估函数中，使定价决策树得到更准确和快速的修剪。</strong></p><p>NFT定价仿若围棋棋局，是看上去简单的复杂游戏，是一个由一连串决策构成的问题。我们可以用同行直觉判断范围，也可以用预言机算法预测未来。</p><p>** 而如果一定要追问什么才是好的定价范式？我认为问题的关键也许就像吴清源说的，不在于目算几遍、目算多远，而在于目算多广、多快和多准。**</p>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
            <enclosure url="https://storage.googleapis.com/papyrus_images/736c664d0700d8f05e91cf7ca1cebe01288d8e54a56224759511ee38c8c5bd6a.jpg" length="0" type="image/jpg"/>
        </item>
        <item>
            <title><![CDATA[上帝在Web3掷骰子吗?]]></title>
            <link>https://paragraph.com/@sallygu/web3-2</link>
            <guid>Qd1mzT2KxfzPDVspLA57</guid>
            <pubDate>Sun, 15 May 2022 17:05:32 GMT</pubDate>
            <description><![CDATA[“We are all fooled by randomness” —— Nicholas Taleb随机 & 命运如果现在有一群极其富有但极其无聊的的比特币巨鲸，决定每年共同拿出1000个比特币举办一次俄罗斯转盘。他们会准备一把左轮手枪，在可装6发子弹的弹夹中只装入一颗子弹，然后对准你的脑袋扣动扳机。 每次扣动扳机所产生的结果可以视为一种平行宇宙。由此在这个单一事件下，总共有对应着6种不同结局的平行宇宙，每种平行宇宙产生的概率相同。 这其中的5个平行宇宙会使你一夜暴富，甚至荣登《Forbes》封面。而剩下的唯一一个平行宇宙则会让你死相难看，贻笑大方。 **那么你愿意参与这个赌局吗？**愿意参与一届？两届？还是每年都来碰碰运气？ 玩这个游戏需要勇气，当然也需要动一些脑筋。83%是一个很高的胜率，确定性达到83%的事情在生活中也并不多见。但如果有些贪婪的傻瓜选择一直玩下去，那20年后他们中的绝大多数应该都可以在报纸的讣闻一栏找到。 在这个假设的俄罗斯转盘游戏里，规则制定得非常清楚，赔率可以计算，风险也可以衡量，所有条件信息都已提前披露。但在现实世界中，我们往往面临的都是非对称信息的...]]></description>
            <content:encoded><![CDATA[<blockquote><p><strong>“We are all fooled by randomness” —— Nicholas Taleb</strong></p></blockquote><h1 id="h-and" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">随机 &amp; 命运</h1><p>如果现在有一群极其富有但极其无聊的的比特币巨鲸，决定每年共同拿出1000个比特币举办一次俄罗斯转盘。他们会准备一把左轮手枪，在可装6发子弹的弹夹中只装入一颗子弹，然后对准你的脑袋扣动扳机。</p><p>每次扣动扳机所产生的结果可以视为一种平行宇宙。由此在这个单一事件下，总共有对应着6种不同结局的平行宇宙，每种平行宇宙产生的概率相同。</p><p>这其中的5个平行宇宙会使你一夜暴富，甚至荣登《Forbes》封面。而剩下的唯一一个平行宇宙则会让你死相难看，贻笑大方。</p><p>**那么你愿意参与这个赌局吗？**愿意参与一届？两届？还是每年都来碰碰运气？</p><p>玩这个游戏需要勇气，当然也需要动一些脑筋。83%是一个很高的胜率，确定性达到83%的事情在生活中也并不多见。但如果有些贪婪的傻瓜选择一直玩下去，那20年后他们中的绝大多数应该都可以在报纸的讣闻一栏找到。</p><p>在这个假设的俄罗斯转盘游戏里，规则制定得非常清楚，赔率可以计算，风险也可以衡量，所有条件信息都已提前披露。但在现实世界中，我们往往面临的都是非对称信息的动态博弈(Asymmetric Information Game Theory)。</p><p><strong>左右我们命运的并不是一把左轮手枪，或是一副扑克牌。我们不知道有多少弹夹中有多少枚子弹，也不知道这副牌有多少花色。在无限多重平行宇宙和无穷大的风险条件下，我们的命运不可预知。</strong></p><p>而这正昭示着我们生活在一个充满随机的世界中。</p><p>如费曼所言，自然界允许我们计算的只是概率。一旦我们具备了计算“概率振幅”的超能力，我们就可以成为先知，预测甚至改变未来。换句话说，<strong>谁掌握了随机性，谁就拿到了命运的骰子，成为了上帝。</strong></p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/84576b55712514e0eaa5f91f45439662215757cd54dd4633ccdd33ef61212be9.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h1 id="h-and" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">随机 &amp; 安全</h1><p>将这种思维平移至加密世界中，生成随机数的机器就可以被视为是传达上帝神谕的使者。也正是因为扮演着这样的角色，预言机(oracle)才由此得名。<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://mp.weixin.qq.com/s?__biz=MzUzNjk5MzI5Nw==&amp;mid=2247489604&amp;idx=1&amp;sn=8f4f50f6ba9a2629d6ccb6dd52c1e7a0&amp;chksm=faec933acd9b1a2c789c096095a9d635a7d5176b97b82005293b9633b98c8dd2434c82cb5c92#rd">过去的文章中</a>，我们已经指出<strong>由于区块链系统的透明性，其链上生成的一切算法、输入、输出都会被暴露给系统参与者。因此在链上直接生成的随机序列实际是可以被预测的。</strong></p><p>一些典型的链上随机数生成方法，比如未来区块哈希，都存在很大的作弊空间和安全漏洞。举一个简单的例子，如果矿工在玩一个博彩游戏，这个游戏的中奖金额是10ETH，而区块奖励的金额却只有3ETH，那么矿工当然有明显动机去作弊。他可以通过丢弃和不发布区块来影响随机序列的生成(Block Withholding Attacks)，从而为自己赢得在游戏中优势。这相当于是<strong>你和矿工打同样一局扑克，但人家可以换底牌。</strong></p><p>而如果将这种可以被篡改和预测的伪随机序列应用到web3各个场景中，我们不难想象其背后的巨大安全风险和不公平性：<strong>一旦黑客或攻击者找到随机算法，并测算出了随机序列的初始输入值，那么无论是NFT的空投发放，GameFi中的极品装备爆出，还是用户个人钱包的私钥地址生成，都有可能被操纵和篡改。</strong></p><p>那究竟怎样的随机数才是真正可信和安全的呢？</p><p>这里我们就需要先对随机数的分类做一个简单了解。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/c5ead4fe3f6bafc30207a3872fa60b77590398e64722efb78050df2a07a03085.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>一般地，我们将随机数分成两种：<strong>真随机数（True Random Number, 后简称TRN)和伪随机数（Pseudo-Random Number，后简称PRN）</strong> 其中，伪随机数又可以被细分为：弱伪随机数和强伪随机数。</p><p>正如 &quot;伪 &quot;这个词所暗示的那样，伪随机数不是真正的随机。所有在区块链上生成的随机数也都是“伪”的。本质上，<strong>PRN是计算机软件根据既定的数学公式和算法指令所输出的看似随机的数字序列，需要外部输入一个初始值来充当seed。</strong></p><p>并且和许多人天然的认知相反，计算机其实很难被用来生成一些偶然的条件，因为它本身就是可预测的设备。元器件是确定的，线路是预设的，代码和算法命令是明确的，在种种固定的外界条件下，我们完全可以通过某些途径（正如上文提到的作弊方式）定位出序列结果。而能够被算出的结果，emm怎么不算……伪随机呢？</p><p>真正的TRN，是从噪音，混乱现象，量子随机过程等物理现象中提取出来随机性，引入计算机才得以生成的。直白一点讲，就是<strong>把链下的一个大骰子接到了计算机。</strong></p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/4767b6e2bd896e72074eb37f9d9d96b0a08a5416e88e52289716f10dc784ac29.png" alt="TRN Mechanism" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">TRN Mechanism</figcaption></figure><p>在性质上，<strong>TRN需要同时具备随机性，不可预测性和不可重现性</strong>三重特质：</p><ul><li><p>随机性：数列完全杂乱无章，不存在统计学偏差，但不代表不会被看穿</p></li><li><p>不可预测性：在知道过去数列的前提下，依然无法预测接下来生成的数列</p></li><li><p>不可重现性：无法重现完全相同的数列，除非将原本的数列保存下来</p></li></ul><p>强PRN只具备前两者，而弱PRN只具备随机性这一个性质，很容易被看穿。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/6a7a9023213681ca6700d660b4b0f591423510ce6dfde913a73e67bd3cf60d7e.png" alt="The Comparison of TRN and PRN" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">The Comparison of TRN and PRN</figcaption></figure><p>我们说一个随机数可信，安全，最起码的就是指它不会被看穿。因此具有不可预测性的TRN和强PRN都可以被用作可信随机数引入智能合约。**弱PRN是不能被应用到比如生成密钥，生成口令密码等等密码学相关的场景中的。</p><p>至于如何生成可信安全的随机数，目前来看最直接高效的方法就是**通过预言机链接外部API并生成加密认证。<strong>比如Chainlink的</strong>可验证随机函数（VRF）<strong>服务，就是通过在测试网络中接入外部API，</strong><a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="http://xn--Random-9p3j357bmxax77dtpi9qfo3ryza484f19eje54xin5jhtxb.org"><strong>比如利用大气现象生成随机数的Random.org</strong></a><strong>，再用公钥和智能合约的seed来验证生成的强PRN，然后将其传输至智能合约。再比如，API3是选择通过和澳大利亚国立大学（ANU）量子光学团队合作，为dApp和Web3各类产品提供</strong>量子随机数生成（QRNG）**服务，将其基于量子现象的生成TRN验证上链交付给请求者。根据其白皮书的相关论述，QRNG能在最大程度上防止女巫攻击，并且这一功能完全免费使用。</p><h1 id="h-and-web3" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">随机 &amp; Web3</h1><p>在理解了可信随机数的定义和来源之后，我们来谈谈一个更实际的问题：随机数在web3到底有什么用？</p><p><strong>事实上，随机数几乎在所有Web3的场景和用例中都不可或缺。</strong></p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/9bcb5ee561f603b6fb5357d0e9cf7864ef039fbe5c12e62aae8deede68da105a.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h2 id="h-gamefi" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">GameFi</h2><p>在游戏中需要用到概率分配得情况比比皆是。典型地，在RPG游戏中，有些角色非常稀有，具有罕见的属性，而有些角色就极为平常。通常越是稀有的角色，就价值越高，越能帮助用户获得高额的经济回报。比如Axie，每个精灵角色都有六个组成部分，每个部分都有不同的属性和能力。游戏厂商想要为用户创造个性化的游戏体验，就必须把角色的分配机制做好，把分配概率设置得尽量公平。</p><p>在PVP对战中，玩家获胜的几率和赢得奖赏的多少都取决于他们在对战中被分配到的对手。如果游戏配对机制不公平，很有可能直接导致玩家的流失。在世界探索类游戏中，玩家在探索地图过程中获得稀有物品和装备的几率也需要一个保证公平的概率模型。在卡牌类游戏中，玩家使用卡牌的顺序很大程度上充当着比赛的胜负手。在射击类游戏中，射击的命中率，角色复活地点的选择都会对玩家的游戏体验造成一定影响。</p><p>上述提到的种种游戏归根结底，本质都是需要<strong>通过注入一定的不可预测性来增强游戏的可玩度</strong>。而不可预测性的产生，就必然需要引入可信的随机数。</p><h2 id="h-nft" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">NFT</h2><p>和GameFi中角色属性具有不同构成一样，NFT的细节也由不同的属性组成。比如BAYC的NFT，就包括背景、衣着、耳饰、眼睛、毛发、帽子和嘴七个不同的属性。而属性的稀有度实际就决定了该NFT的价值。对于稀有度分配和属性总量上限的制定，都可以通过利用可信随机数来解决。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/05714eda0dcc0c0e3d83f9d2b89ebc0d7891530081f9a85c672fa80dc5c489d9.png" alt="BAYC Property Categories" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">BAYC Property Categories</figcaption></figure><p>此外，随着越来越多NFT社区的建立和成长，项目方也经常会在达成某些里程碑时或是特殊节日里，通过发放随机的NFT空投奖励来激发成员的忠诚和活跃度。而空投哪些地址，怎样保证空投的公平性，都是需要项目方谨慎地根据一定发放机制来执行。可信随机数在这类情景下就可以被视为一种良好的解决方案。事实上如今大部分NFT的空投激励活动也都是集成Chainlink VRF等服务开展的。</p><h2 id="h-defi" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">DeFi</h2><p>类似Olympus这样的DeFi协议和算法稳定币协议中，其质押机制通常会设有变基质押（rebase）流程来解决份额稀释问题，或是激励特定的市场行为，比如保持算稳的价格锚定。而rebase机制一个“敏感性”很强的变量就是rebase的周期时间。设定的rebase周期时间如果长期不变，就很可能会被短期套利者利用。因此rebase机制设计可以应用可信随机数，来决定一个变更调整的大致周期。 在一些创新型的Yield Farming协议中，其奖励机制也会引入一定的波动性来提高用户的参与度。参与度高贡献大的用户的年化收益率会区别于普通或是参与度较低的用户。因而资金池中奖励大小的分配和收益幅度的增减，都会需要使用随机数来调节和建立模式。</p><h2 id="h-dao" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">DAO</h2><p>在DAO的治理过程中经常需要成员为组织做出一些关键决策。而一个庞大的DAO社区，往往可能通过设立董事会等核心管理层来更高效地执行日常决策。而这些董事会和管理人员的选择和增补，会需要一个公平透明的机制来进行决策。引入可信随机数对于这类机制的设立可以被认为是一个最好和最便捷的选择。 此外，在一些涉及到大量资金管理或是专门进行投资的DAO中，随机数也可以被用来提高DAO成员身份认证系统的安全性，防止黑客或恶意攻击者逆向工程身份认证过程，保证DAO内资金库的安全。</p><h2 id="h-l1vrf" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">L1共识机制中VRF的运用</h2><p>新型公链需要面对的一个很大问题就是分片（shard）后安全性的降低。在未分片前，网络中所有的验证者共同验证所有交易，但在分片后，这些验证者被平均分配到每个分片上的进行验证，这就使得每个分片被fork（51%攻击）的几率大大提升。解决这个安全性问题的一个良好方案就是让攻击者无法提前知道每个分片的验证者是谁，而这就需要引入一定的随机性和不可预测性。NEAR就是通过VRF服务对验证者进行随机分配，从而隐藏验证者具体验证的分片信息。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/416e35e561f9f2ffde0121f820a376e592c4245f8549bf73474a81ecac2782cb.png" alt="Source: NEAR Nightshard protoco" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">Source: NEAR Nightshard protoco</figcaption></figure><h1 id="h-" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">写在最后</h1><p>随机性之所以令人着迷，是因为它意味着不确定，意味着一种可能性。可能性能为我们带来希望，也会为我们带来恐惧。所以决定论与非决定论的争执绵延至今，而上帝究竟掷不掷骰子其实也没有那么重要。</p><p>也许世间万物都在无可挽回地走向大数定律，也许所有宿命般的相识都不过是概率上的偶然，也许你、我、他都只是在Web3随机漫步的傻瓜。</p><p>但不知道答案，总是比得到一个错误的答案有意思的多。正如卡洛·罗威利在书中提到的那样：<strong>“What’s nonapparent is much vaster than what’s apparent”。</strong></p><p><strong>保持愚蠢，敬畏随机。</strong></p>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
            <enclosure url="https://storage.googleapis.com/papyrus_images/615e1e562a34748751143e24584e34e7829818ee6ff82dbb22cf2f2a669da326.jpg" length="0" type="image/jpg"/>
        </item>
        <item>
            <title><![CDATA[A Random Walk Down VC (03) - Benchmark]]></title>
            <link>https://paragraph.com/@sallygu/a-random-walk-down-vc-03-benchmark</link>
            <guid>7duYdpgHFqK1D9HyepX3</guid>
            <pubDate>Wed, 04 May 2022 14:54:57 GMT</pubDate>
            <description><![CDATA[Billy GurleyBill Gurley has spent 18 years as a GP at Benchmark. He has invested in and served on the board of such companies as Jamdat, GrubHub, Nextdoor, OpenTable, Stitch Fix, Uber, and Zillow.com. Before entering the VC business, Bill spent four years on Wall Street as a research analyst. His research coverage included such companies as Dell, Compaq, and Microsoft, and he was the lead analyst on the Amazon IPO.Who can be VCsThe truth is anyone can be a VC because handing out money is one ...]]></description>
            <content:encoded><![CDATA[<blockquote><h1 id="h-billy-gurley" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Billy Gurley</h1><p>Bill Gurley has spent 18 years as a GP at Benchmark. He has invested in and served on the board of such companies as Jamdat, GrubHub, Nextdoor, OpenTable, Stitch Fix, Uber, and Zillow.com. Before entering the VC business, Bill spent four years on Wall Street as a research analyst. His research coverage included such companies as Dell, Compaq, and Microsoft, and he was the lead analyst on the Amazon IPO.</p></blockquote><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/5d9b188fd1725e16310bef1a8a12d408a22649e9ad5cd5326720b58bf4e22a03.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h1 id="h-who-can-be-vcs" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Who can be VCs</h1><p>The truth is anyone can be a VC because handing out money is one of the lowest barriers to entry jobs. </p><p>The real question is - <strong>can you be successful?</strong></p><p>Prior to Benchmark, my first venture job was with a firm called Hummer Winblad and one of the co-founders was an athlete playing in the leagues for six years. So athletes can also have a successful career in VC.</p><h1 id="h-how-to-make-an-investment-decision" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">How to make an investment decision</h1><p>You need to know how to judge the qualities of a good leader or someone who is a team player. There are many times we made an investment solely because of the following two points:</p><ol><li><p><strong>The power of founders who dares to make a bold move</strong> like business transformation. Discord, Nextdoor, Docker and Slack are all pivots, which means the company stopped doing what they were doing at first and started doing something different. </p></li><li><p>The <strong>GTM capabilities</strong> of founders and this quality is somehow missed by a lot of people. Unit economics really matters. If you look at the most successful companies, they weren’t always the first but they were the ones that <strong>figured out how to get to market fastest.</strong></p></li></ol><p>I recommend five books if you’re gonna be investing:</p><ul><li><p>Michel Porter’s Competitive Strategy</p></li><li><p>Crossing the Chasm</p></li><li><p>The Innovator’s Dilemma</p></li><li><p>Shoe Dog</p></li><li><p>Jerry Kaplan’s Starup</p></li></ul><br><h1 id="h-why-invest-uber" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Why invest Uber</h1><p>We had made money on Open Table, and had a belief that an <strong>ordering network</strong> can be put on top of multiple industries. </p><p>We were particularly intrigued with black cars and had met with a bunch of taxi startups that were putting this model on taxis. But taxis are usually an oligopoly market and the price is fixed. Thus we felt that would not work well and looked for someone doing this on black car.</p><p>I actually referenced Travis to Mark Cuban who had invested in his other startup previously. Though the previous one failed, we all thought this time <strong>the product market fit was obvious</strong> and he was not gonna let it fail.</p><h1 id="h-investment-advices" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Investment Advices</h1><p>My favorite Warren Buffett quote is “<strong>There is a fool in every market and if you don’t know who it is it’s probably you</strong>”.</p><p>If we see something pop or somebody makes billions of dollars but it was in a category we don’t understand we try not to have anxiety about that because we want to invest where we believe and we have a competitive advantage.</p><h1 id="h-why-benchmark-successful" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Why Benchmark successful</h1><p>Our founding partners realize that in a lot of professional firms, the senior people take all the economics while the junior people do all the work, so they want to find a way where that wouldn’t be true. </p><p>We created a concept of an <strong>equal partnership</strong> where the firm is lean and small, but all the investing professionals make the exact same compensation. We believe that it <strong>generates a phenomenal amount of teamwork</strong> because other firms with differential pay, every time they raise funds their members are holding each other’s throat , competing with each other. And it also allows us to <strong>recruit the very best people</strong> we want to recruit since we’re offering someone an equal seat at the table.</p><p>We only do early stage, so if you want to write a hundred millions dollar check to mature companies you need a huge set of analysts to pour through all the due diligence. But in the early stage, it’s two people on a PowerPoint, it’s a lot more about the individuals than anything else. </p><h1 id="h-investment-dilemma" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Investment dilemma</h1><p>We are in a remarkably unique time of the history of business where <strong>interest rates have been low for 5-7 years</strong> and it’s never happened before. Some of the best investors in the world like Buffett and Howard Marks, are complaining about this. If you set interest super low it increases the amount of speculation people are willing to do because there’s no yield anywhere else.</p><p>Stanley Druckenmiller was quoted as saying “I got nothing against Taxens, but if you let them drill holes with other people’s money, they&apos;ll drill lots of holes.”</p><p>The quintessential contribution to this getting more profound was when <strong>Softbank raised the Vision Fund and started writing checks into the billion-dollar sizes</strong> into these private companies and it creates a prison’s dilemma and the hardest problem of all time in making decisions because you have such an ambitious competitor.</p><p>Also, Wall Street is now asking a lot of questions about unit economics. When Amazon first came public, it was losing money but <strong>eventually it had to get profitable and the street had to be convinced</strong>, so it will be an evolution as you move into being public and being more scrubbed by Wall Street.</p><h1 id="h-how-you-predict-the-future" class="text-4xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">How you predict the future</h1><p>One of my partners has a saying that “<strong>We’re not paid to see the future, we’re paid to see the present very clearly</strong>.”</p><p>It’s very obvious to me that data is gonna be ever more present all the time. But I wonder how much more information is going to be available five years from today and in what ways can we expect that and what business opportunities will that create.</p><p>The <strong>geofencing, geolocation</strong> tool inside of apps, every week I see an entrepreneur doing something new with that and it’s really interesting. There is still a lot of greenfield space there.</p><p>The open-source is blowing up, SaaS is blowing up, security spending is blowing up, those seem like never-ending  right now.</p>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
            <enclosure url="https://storage.googleapis.com/papyrus_images/b5e391d2bfa8349d404999440ae70358dc2cd01a0dd7ad5898da29f6decd9ba2.png" length="0" type="image/png"/>
        </item>
        <item>
            <title><![CDATA[A Random Walk Down VC (02) - Andreessen Horowitz]]></title>
            <link>https://paragraph.com/@sallygu/a-random-walk-down-vc-02-andreessen-horowitz</link>
            <guid>csRh24lVPvUlKD3HozBD</guid>
            <pubDate>Sun, 01 May 2022 14:16:22 GMT</pubDate>
            <description><![CDATA[Mark AndreeseenAndreessen is a co-founder and general partner at Andreessen Horowitz, and has invested in companies such as Facebook, Pinterest, Twitter and Skype, among others. He co-created the highly influential Mosaic internet browser and co-founded Netscape, and has been named one of the 100 most influential people in the world by Time.Two types of founder-CEO model1.Mark Zuckerburg+Sheryl model 2.Bill Campbell+Scott Cook model (Intuit) / Dick Costolo model (twitter)/ Steve Jobs +John Sc...]]></description>
            <content:encoded><![CDATA[<blockquote><h2 id="h-mark-andreeseen" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Mark Andreeseen</h2><p>Andreessen is a co-founder and general partner at Andreessen Horowitz, and has invested in companies such as Facebook, Pinterest, Twitter and Skype, among others. He co-created the highly influential Mosaic internet browser and co-founded Netscape, and has been named one of the 100 most influential people in the world by <em>Time.</em></p></blockquote><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/3c3a0b25c0a08a90cd9907aaed56f2850d0659c09876ba3a11a38820036e67e5.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h2 id="h-two-types-of-founder-ceo-model" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Two types of founder-CEO model</h2><p>1.Mark Zuckerburg+Sheryl model</p><p>2.Bill Campbell+Scott Cook model (Intuit) / Dick Costolo model (twitter)/ Steve Jobs +John Scrulley model (apple)</p><p>The founders don’t have the capabilities to be a CEO. They hire a business person to be the CEO and the CEO understands the role of founders, product and tech strategy more. For example, Bill Campbell is not a tech or product person, but is an outstanding operator with profoundly deep respect for founders, tech and product.</p><h2 id="h-post-investment-service-model" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Post-investment service model</h2><p>a16z builds a disruptive model, providing value-added services, including <strong>hiring and marketing.</strong></p><p>Many VC firms have tilted hard in the direction of <strong>professional sales driven CEO.</strong></p><p>a16z partners &amp; GPs almost all have founders/CEO experience and it is necessary to get somebody on the board who really has been through the war, really understands what things are like.</p><p>So when things went wrong, key engineer quits, the founders can’t get along, the biggest customer dumps you, competitor comes out with a much better product, can’t raise money……<strong>need an advisor in the board can say: I was in that situation before and I can give you some advices on what does work and what doesn’t.</strong></p><h2 id="h-why-should-a-tech-founder-become-a-pro-ceo" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Why should a tech founder become a pro-CEO</h2><p>You can have an <strong>enormous network of executives</strong>, know all the recruiters, reporters and editors, VCs and technical founders often don’t because they have been heads down, writing codes most of their lives. So we can inject him into the network that gives him the superpowers of the network that is comparable to what John Chambers (ex Cisco CEO) might have.</p><p>a16z has about <strong>60 full time professionals</strong>, across five operating teams in the firm that are not GPs.</p><p>For example, when it comes to finding a Sheryl, we have built very deep relationships with all the Sheryls and will have very easy access to them.</p><h2 id="h-traits-of-founders-a16z-values" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Traits of founders a16z values</h2><p>Basic math component: about 4000 startups a year found in the tech industry want to raise capital, a16z can invest about 20, the fall off is significant.</p><p>The hard thing is deciding which ones we are going to invest in. About 200 deals a year can get funding by top VCs. Among the 200, 15 of those will generate 95%+ of all the economic return.</p><p><strong>Even the top VCs will generally lose half of their deals.</strong> So even funded by top VC in not complete validation.</p><p>The thing all the top VC firms have in common is they did not invest in most of the great successful tech companies, which is incredibly frustrating.</p><p><strong>The decision ultimately is around people, as like 90% of the decision.</strong></p><p>The two things we really zero in people are <strong>courage and genius.</strong> Courage is the one people can learn, namely not giving up in the face of adversity, being absolutely determined to succeed. You can force yourself to do it. The genius part is hard to force yourself to do.</p><p>Courage without genius might not get you where you need to go, but genius without courage almost certainly won’t.</p><p>We are looking for some kind of magic combination of genius and courage. We are strongly biased towards <strong>people who are so determined to succeed that they just never give up.</strong></p><h2 id="h-two-kinds-of-vc" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Two kinds of VC</h2><p>1.<strong>Top-down thinking</strong> <strong>approach</strong>: Sequoia, Accel, Bessemer, Kleiner Perkins</p><p>Very explicit about how they think about products and markets. Run an annual planning process at the beginning of the year and draw a value chain map of what they think markets are gonna like.</p><p>For example: Networking: fiber, optics, communication chips——routers——ISPs——wireless business</p><p>List boxes for each of the product categories. VCs can invest in one company per category and the goal for the year is to put a name in each box. At the end of the year, if they had invested in the best possible company in each box they identified.</p><p>2.<strong>Benchmark approach</strong>, Arthur Rock approach: a16z</p><p>thought the entire art of VC is the big breakthrough ideas, and the nature of ideas is unpredictable. <strong>The really breakthrough ideas often seem hard to be understood at the first time and the fact they seem nuts can be a very positive signal because</strong>：</p><p>1.explains why the thing is not being done by existing big companies, it’s too strange,</p><p>2.if it works, the bit flips at some point, then you know those are companies that can just explode and are gigantically huge.</p><p>Bitcoin, Airbnb, Uber all seem nuts. We want to kinda tilt into the really radical ideas but it can’t be predicted so you just learn as much as you can, enter as close to a zen like blank slate as the beginning.</p><p>We are trying really hard to let ourselves <strong>be educated by the really smart entrepreneurs.</strong></p><h2 id="h-preferred-pitch" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Preferred Pitch</h2><p>The worst pitch meeting: snapchat for dogs, airbnb for parents, infinite variations of all the successful ones (copycats)</p><p>The best: really bright founder who has done a tremendous amount of work and completely understands the domain, and walks in with a really crazy idea and in the course of an hour can basically walk you through “the idea of maze”.</p><p><strong>The entrepreneurs who really have the radical ideas are generally not only uncoachable and generally react with hostility to being coached.</strong></p><p>We will do a test:</p><ul><li><p>Q：“Hey, how about doing this in another approach?”</p></li><li><p>A1: staring at you like you idiot, moron and explain in detail why you are wrong √</p></li><li><p>A2: “Oh, that’s a great idea!“ ×</p></li></ul>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
            <enclosure url="https://storage.googleapis.com/papyrus_images/def7cc842fb91f3c87a6370703996026d090717f6fe4140704acf0776d642485.png" length="0" type="image/png"/>
        </item>
        <item>
            <title><![CDATA[A Random Walk Down VC (01) - Sequoia]]></title>
            <link>https://paragraph.com/@sallygu/a-random-walk-down-vc-01-sequoia</link>
            <guid>wkOotX5w6Wxz2wb3sO5L</guid>
            <pubDate>Sun, 01 May 2022 13:35:47 GMT</pubDate>
            <description><![CDATA[Don Valentine: Gandfather of VCDon Valentine founded Sequoia in 1972, before the terms “Silicon Valley” and “venture capital” had been coined. His original investments includes Apple, Atari, Oracle, Cisco, etc. Prior to founding Sequoia Capital, Don served as a sales and marketing executive at Fairchild Semiconductor for seven years before he went on to establish National Semiconductor. Don obtained his Bachelor of Science in Chemistry from Fordham University in 1954.Why sequoia successfulWe ...]]></description>
            <content:encoded><![CDATA[<blockquote><h2 id="h-don-valentine-gandfather-of-vc" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Don Valentine: Gandfather of VC</h2><p>Don Valentine founded Sequoia in 1972, before the terms “Silicon Valley” and “venture capital” had been coined. His original investments includes Apple, Atari, Oracle, Cisco, etc. Prior to founding Sequoia Capital, Don served as a sales and marketing executive at Fairchild Semiconductor for seven years before he went on to establish National Semiconductor. Don obtained his Bachelor of Science in Chemistry from Fordham University in 1954.</p></blockquote><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/06edb618588e3a5e96cbc314a77298a5b4b90074ce0c8669f5302b01200196b5.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><h2 id="h-why-sequoia-successful" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Why sequoia successful</h2><p>We focus on the size of the market, the dynamics of the market, the nature of competition because our object is always to build big companies.</p><p><strong>We don’t choose people, we choose markets.</strong> We don&apos;t care about the background of the founders and we rarely ever invest in an area with only one product.</p><p>Take Apple for example, we finance more than one memory company or disk drive company. Without the memory system, the PC is nothing. Apple found sth. they are needed in Xerox PARC and we finance all these things. This is how we choose to invest on a system application level. We looked at the system and made over 15 investments in the category.</p><p>We also invested in many <strong>inexperienced young people and taught them outsourcing.</strong></p><p>The only thing that matters is they must be very good at <strong>tech and engineering</strong> and that normally took us about six people to start the company. The secondary functional skill we are interested in is <strong>marketing</strong>.</p><p>We think creating a market is too expensive for our checkbook and are more interested in exploring markets early.</p><p><strong>Good people are technologists with no interest in becoming wealthy</strong> and keen on solving tech problems and creating new products.</p><p>Unlike most VC people, we don’t wait for you to knock on our door, we knock on your door. We are actively looking for the right people with the capability to solve the problem we are interested in.</p><p>We have a team with different expertise in developing a knowledge base  and market sizing info. We tend to invest application system approach.</p><h2 id="h-why-storytelling-critical" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Why storytelling critical</h2><p>The art of storytelling is incredibly important because money flows as a function of the stories.</p><p><strong>Learning how to ask a question</strong> is way more important than anything in the world (20 words maximum). For many investments we made, we didn&apos;t understand the answers, but we constantly worked on developing the questions.</p><p>When a company we invested in failed, we always have <strong>post mortems</strong> to figure out what we missed, what questions we didn’t ask.</p><h2 id="h-which-people-we-are-looking-for" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Which people we are looking for</h2><p>We are less interested in their education than in <strong>what they did in their prior companies.</strong> We make the management system as easy as possible and if we can outsource something, never make it in-house.</p><p>Only one metric in finance matters to us and that is cash flow. We hire people who are wizards at <strong>cash flow.</strong></p><p>We don’t need balance sheets and we love recessions.</p><h2 id="h-lessons-from-worst-investment" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0">Lessons from worst investment</h2><p>What didn&apos;t work is the dynamics of markets. We developed some spectacular things for which there were no buyers. The critical thing is  getting a product developed where the timing of the product’s availability and the market demand are simultaneous.</p><p><strong>We didn’t finance our portfolio companies any more if no one wanted their products.</strong></p><p>But when those companies meet the milestones, we then refinance them.</p>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
            <enclosure url="https://storage.googleapis.com/papyrus_images/ea02115284c603c11577403fbcee866205c27b55f88de787b28d766f40e8033d.jpg" length="0" type="image/jpg"/>
        </item>
        <item>
            <title><![CDATA[Solidly: 新游戏，老玩家]]></title>
            <link>https://paragraph.com/@sallygu/solidly</link>
            <guid>c1onFoXulOJ9pgUoBSh2</guid>
            <pubDate>Tue, 19 Apr 2022 03:37:01 GMT</pubDate>
            <description><![CDATA[注：本文始发于2022年2月24日，后续AC退圈带来的负面影响并未包含在本研究时段内，仅供机制研究参考。长久以来，传统的AMM模式一个很大的问题就在于协议平台和LP之间的冲突。因为LP的收益率实际上和所提供的流动性资金量大小并不成正比，协议所发放代币市场价值的缩水，和交易对在市场波动中形成的无常损失，都会使LP蒙受“塌矿”的风险。而为了解决这种协议和市场参与者在利益上此消彼长的零和局面，Yearn的创始人AC(Andrew Cronje)联合知名加密社区Frog Nation的创始人Dani(Daniele Sestagalli)推出了名为“Solidly”的全新DeFi2.0协议项目，并迅速在加密社区引发了讨论狂潮。 截至今日Solidly的TVL已近4亿美元，24小时成交量涨幅超1300%。随着越来越多的目光聚向这个全新协议，我们将通过本文，对其从协议机制和引发的吸血鬼战争两方面进行解构。TVL Volume (Feb 2022)Solidly是什么：ve+(3,3) Solidly的原身是基于AC创建的一个新型DEX类项目ve(3,3)。而ve(3,3)，正如其名称一样，简...]]></description>
            <content:encoded><![CDATA[<p>注：本文始发于2022年2月24日，后续AC退圈带来的负面影响并未包含在本研究时段内，仅供机制研究参考。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/2039222800147107b85c328ab82400991de3878b6e07ecd735bc92d1beb080e1.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>长久以来，传统的AMM模式一个很大的问题就在于协议平台和LP之间的冲突。因为LP的收益率实际上和所提供的流动性资金量大小并不成正比，协议所发放代币市场价值的缩水，和交易对在市场波动中形成的无常损失，都会使LP蒙受“塌矿”的风险。而为了解决这种协议和市场参与者在利益上此消彼长的零和局面，Yearn的创始人AC(Andrew Cronje)联合知名加密社区Frog Nation的创始人Dani(Daniele Sestagalli)推出了名为“Solidly”的全新DeFi2.0协议项目，并迅速在加密社区引发了讨论狂潮。</p><p><strong>截至今日Solidly的TVL已近4亿美元，24小时成交量涨幅超1300%。随着越来越多的目光聚向这个全新协议，我们将通过本文，对其从协议机制</strong>和引发的<strong>吸血鬼战争</strong>两方面进行解构。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/db240e12c662c7b3004c5c3b779130fbfa8eeb5d18c4e01644521d3668bdaf29.png" alt="TVL Volume (Feb 2022)" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">TVL Volume (Feb 2022)</figcaption></figure><p><strong>Solidly是什么：ve+(3,3)</strong></p><p>Solidly的原身是基于AC创建的一个新型DEX类项目ve(3,3)。而ve(3,3)，正如其名称一样，简单来说就是一个基于Curve的veCRV经济架构，并融合Olympus的(3,3)博弈理论，优化后所集成的一个代币发放机制。** **</p><p><strong>veCRV: 投票&lt;&gt;流动性</strong></p><p>veCRV是Curve创立的一种新型代币经济模型，想要了解它必须先对Curve有一个基础的认知。Curve是一个采用AMM模式主要针对大规模稳定币的DEX，因此需要LP提供交易对。作为回报，Curve会将自己的CRV代币和平台50%的交易手续费作为对LP的奖励。</p><p>LP在拿到CRV后，可以选择将代币质押到平台内锁仓（最多可以锁仓4年），而平台就会返还veCRV（ve: voting escrow 投票托管）作为权益证明，锁仓时间越长，veCRV返还的就越多。我们可以把质押这个操作当成是对Curve股权的购买，持有veCRV就意味着持有了股权。<strong>长时间的锁仓即长拿股权，代表着对平台的一种深度信任。平台自然也愿意为钻石手的忠实股东，而非只想“挖提卖”的短期投机者，发放更多收益。</strong></p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/2341e9efa1cee39da3b46cd1e9a619c24fd6598ce2fb56e53b2ea7c55e3415e2.png" alt="veCRV Return Mechanism" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">veCRV Return Mechanism</figcaption></figure><p><strong>veCRV的锁仓回报机制</strong></p><p>同时，持有veCRV的股东也拥有对池子中奖励发放份额投票的权利。这种设计正是ve模型中最令人称道的一点：**以投票激励流动性。**我们常听到的所谓Curve war实际上就是对分配投票权力的争夺。因为veCRV的持有者可以通过投票，帮自己的池子拿到更多奖励发放的份额。而奖励份额的增多必然导致更多LP的涌入，这样背靠充足的LP，池子的交易体验自然也越好，交易手续费随之水涨船高。在这样的正循环下，veCRV的持有者就可以持续获益。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/21bfbef208c01e76cbe4897c3b37fa5cdc5f2900876dd295d1250c241ad4bd07.png" alt="veCRV Value-add Matrix" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">veCRV Value-add Matrix</figcaption></figure><p><strong>veCRV机制目前的缺陷</strong></p><p>然而，veCRV的模式也并不是十全十美。</p><p>一方面，CRV的持续发放，会使得用户的股权被不断稀释。随着股权占比的不断减小，veCRV持有者所能分得的价值实际上会逐渐缩水。</p><p>另一方面，由于veCRV直接绑定了LP地址，这种权益证明是不可交易或转让的。作为奖励的50%交易手续费会按比例直接打给veCRV持有者。因此这在很大程度上限制了流动性。</p><p><strong>(3,3): 锁仓→纳什均衡</strong></p><p>(3,3)是DeFi2.0协议组织Olympus DAO将博弈论引入其算法稳定币OHM后所提出的一种纳什均衡理念。在进入对(3,3)博弈论的剖析之前，我们首先有必要对OHM的激励机制一个简单梳理。OHM的激励主要有两种方式：一种为债券机制(Bond)，另一种为质押(Stake)机制。</p><p>在债券机制中，用户可以通过把一些价值币如BTC，ETH等以LP的形式交给Olympus协议，从而<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://web.stanford.edu/~guillean/papers/ohm-staking.pdf">以低于市价Pi(t)的折扣价Pd(t)拿到OHM</a>。这种债券发放会分成5天，所以债券便宜的原因就和某些网购邮费便宜的原因一样——配送慢，你要多等一会才能签收想要的东西。</p><p>而在质押机制中，用户可以直接在市场上用价值币兑换OHM，再把OHM质押到平台上。Olympus拿到用户手上的价值币后会将其存入国库，而一个OHM的真实价值就等于国库中价值币总值除以OHM的总发行量。若一个真实OHM的价值是1USDT，而市场现价高于了这个点，协议就会增发OHM以平衡价格，并把这些增发的OHM按比例打给质押用户，这就解决了veCRV模型中所存在的份额稀释问题。<strong>因此根据Olympus这套所谓的变基质押(rebase)流程，理想状态下只要用户都选择长期质押，其仓位内的OHM余额就可以不断复利增长。然而如果二级市场上的OHM抛压严重，这种飞轮就无法持续。</strong></p><p>综上，Olympus提出的(3,3)纳什均衡正是建立在这两种激励范式上。根据传统经济学理论，纳什均衡指的是在非合作博弈中，两个及以上参与者达成共同利益最大化的一种策略。德扑爱好者也可以将其理解为德扑中的GTO（Game Theory Optimal）。根据Olympus的官方文件，假设市场上存在两个参与者，则双方的三种行为将对应着以下9种结果：</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/765aeaea2f0952adc94f739b50bbe4ac514000e83fc00e4ddc53d373d34067ba.png" alt="Olympus Game Theory Matrix" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">Olympus Game Theory Matrix</figcaption></figure><p>根据这个矩阵，我们可以看到质押和折扣债券都可以为协议带来正向收益，而双向质押即为可以实现纳什均衡的最优策略。相反，若参与者都持对立立场选择卖出，那这场博弈中两方都会成为输家。但<strong>事实上不管是在DeFi还是在牌桌，所有人的终极目标都是为了赢，双赢说到底也是为了自己赢。Olympus引入这套机制也是希望倡导合作实现利益最大化的共赢理念。</strong></p><p>** **</p><p><strong>Solidly的创新：veNFT+投票权优化</strong></p><p>Solidly深度缝合了以上所述veCRV和(3,3)模型的特点。一方面，Solidly采用ve机制返还veROCK（由发放的ROCK代币引申）作为股权的证明。另一方面，Solidly引入(3,3)变基质押范式，使得质押者股权不会因时间推移而被稀释。但在这个基础上，Solidly也做出了自己的几点创新：</p><ol><li><p>**将投票权转换为NFT，解放流动性。**意识到了用户在质押后无法转移veCRV的限制，Solidly将返还的veROCK设计成了NFT的形式，解放了用户在二级市场自由交易的权利。质押后，Solidly的用户可以把代表股权的NFT转让给其他人，任何NFT的持有人都有投票权来决定奖励的分配。</p></li><li><p>**质押者得到100%交易手续费，但只能从投过票的池子赚取奖励。**和Curve中所有veCRV持有者都能平分一半手续费不同，veROCK持有者可以赚取全部的交易手续费，但仅限在自己投过票的池子中。所以在这种模式下，投票的动作就至关重要。理想的状况下，所有投票参与者都会为实现自身利益最大化，将票投给能赚取最多手续费的池子，使这些池子获得最多的奖励分配。而一旦质押者投错了池子，那他将得不到任何奖励或者只能得到很低的费用。</p></li><li><p>**代币分发量根据TVL比例调整。**在Solidly的模式下，发放的代币数量和总质押数量呈反比。假设协议最初有5000流通代币，一周最多发放500奖励代币，我们可以分成三种情况进行讨论：</p><ol><li><p>在总质押数为0时，500奖励代币都被发放出去。这种情况的LP完全不在乎手续费的收入，只想挖矿套现，往往出现在LP对协议失去信任不愿持有股权的时候。</p></li><li><p>当5000代币都被质押时，奖励代币就不会被发放。这种情况下协议可以用手续费的收入来支撑整个系统，所以对协议来说，最佳方案就是代币被尽可能地质押，避免发放奖励代币。</p></li><li><p>在总质押数为2500时，50%的奖励代币即250枚代币被发放出去。由于个人和协议利益不可能完全重合，所以实际上这种情况更为常见，大部分时间质押比例和代币发放会在一定区间内呈现出一种动态平衡。</p></li></ol></li></ol><p>** **</p><p><strong>空投引爆的吸血鬼战争</strong></p><p>**比Solidly的底层协议机制更为人津津乐道的，是其所引发的吸血鬼战争（Vampire war）。**1月6日AC在twitter宣称Solidly代币ROCK的发放将直接空投给fantom协议上的前20名锁仓量最大的协议（最后空投了前25的协议）。在得到空投代币后，这些协议可以自行决定如何使用，可以选择把veROCK分配给社区，也可以选择用来激励交易池，为他们自己的代币赚取回报。而这短短一句话，直接引爆了Fantom链上协议间的吸血鬼攻击（Vampire Attack）并逐渐演变为大战。仅仅是为了争夺TVL，挤进前20拿到代币空投，**两个DAO——0xDAO和veDAO就此应运而生。 **</p><p><strong>VeDAO：内卷先锋</strong></p><p>VeDAO在1月19日由匿名开发者发起，组建的目的非常纯粹，就是为了吸引资金抢占TVL，以满足空头奖励快照的要求。在发起的短短一天内，VeDAO就通过发放代币集资了超过10亿美元，在Fantom上的TVL一度跃升至第二位。VeDAO的整个网页相当简单，只提供一种复刻Sushi MasterChef流动性挖矿的功能。用户可以这种流动性挖矿，来获得WEVE代币参与DAO 的治理。而VeDAO则通过用WEVE换取来的稳定币在短时间内积累提升自己的TVL，并许诺所有DAO成员都可以在将来共同管理veROCK并决定其分配去向。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/da301c72f65f58b0904fee941bffe7575679ae623d8bfd0fb57c305071bed5a6.png" alt="veDAO homepage" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">veDAO homepage</figcaption></figure><p><strong>0xDAO：复仇者联盟</strong></p><p>VeDAO的发轫毫无疑问损害了Fantom链上其他协议的利益。在受到VeDAO肆无忌惮的偷猎后，许多Fantom OG和原生协议决定开始反击。五个Fantom的优质原生协议spookyswap，scream，liquid driver，hundred Finance，RevenantFinance联手组建了复仇者联盟0xDAO进行反击。0xDAO模仿VeDAO的策略，承诺将其所有代币OXD都分发给LP，持有OXD就可以参与决策veROCK的出售和分红，并成功短短几天内就集资到了20亿美金的TVL。最终，0xDAO也不负众望赢得了战争，以接近12%的链上TVL锁仓比拿到了最大的空投份额。</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/d2a4d48be2161fe4566b72b97153bb9cf8e556599d5dc423345d5646468c17d9.png" alt="完整空投名单及TVL(ve(3,3): An Introduction into the New and Ambitious Solidly Exchange (substack.com))" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">完整空投名单及TVL(ve(3,3): An Introduction into the New and Ambitious Solidly Exchange (substack.com))</figcaption></figure><p>值得思考的是，Solidly的未来将会走向何方？在对Curve和Olympus DAO的经济模型进行了创新，进一步释放了流动性的同时，我们也看到了其目前发展存在的隐患：</p><p>**1.FTM链上硝烟又起。**在抢夺TVL的吸血鬼战争结束后，以0xDAO和Solidex为首的空投赢家又将成为Fantom链上的Convex设为目标，开始酝酿新一轮更为激烈的veROCK战争。</p><p>**2.协议创新无法挣脱囚徒陷阱。**尽管Solidly在(3,3)的博弈理论上进行了优化，但人性依然很难经得起考验。一旦DAO的的治理惩罚机制失效，参与者们往往会因猜疑和恐惧而放弃共赢，<strong>一个背叛者的抛售就足以形成多米诺骨牌效应，使整个市场陷入囚徒困境(Prisoner&apos;s Dilemma)的倾轧和霍布斯口中“所有人对所有人的战争“中。</strong></p><p>**3.veROCK被赋予的价值还非常有限。**虽然其可以决定未来奖励发放份额，但也只是一个相对基础的功能。如果仅仅是永续地释放这些代币，协议在运营上很可能无法持续。在2月10日上线后，Solidly已宣称其DAO正在研究讨论这一话题。</p><p>究竟Solidly是照耀defi2.0的明灯，还是一场包装精致的庞氏骗局？我们将对后续的发展继续保持关注。</p>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
        </item>
        <item>
            <title><![CDATA[How does oracle make Web3 a better place?]]></title>
            <link>https://paragraph.com/@sallygu/how-does-oracle-make-web3-a-better-place</link>
            <guid>f5s6sJoCAZtSRkn2sMlJ</guid>
            <pubDate>Tue, 19 Apr 2022 03:05:44 GMT</pubDate>
            <description><![CDATA[Oracle：trust builderOracle is usually reckoned as the bridge and window of on-chain and off-chain data. In short, oracle is a middleware providing real-world data services for blockchain projects.Source: IOSG If we recognize the definition that blockchain is the trust machine, then oracle can be essentially referred to as the** trust-maintenance machine.** As a matter of fact, the trust generated by blockchain itself is usually not sufficient to support all the demands of upper layer applicat...]]></description>
            <content:encoded><![CDATA[<h2 id="h-oracletrust-builder" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>Oracle：trust builder</strong></h2><p>Oracle is usually reckoned as the bridge and window of on-chain and off-chain data. In short, oracle is a middleware providing real-world data services for blockchain projects.</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/92eb1ff5d6d8cb3735b637fd8923b1977271ab812979fbef854049ba92fa3623.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>Source: IOSG</p><p>If we recognize the definition that blockchain is<a target="_blank" rel="noopener noreferrer nofollow ugc" class="dont-break-out" href="https://www.economist.com/leaders/2015/10/31/the-trust-machine"> the trust machine</a>, then oracle can be essentially referred to as the** trust-maintenance machine.** As a matter of fact, the trust generated by blockchain itself is usually not sufficient to support all the demands of upper layer applications. More trust is needed to be fed in from the oracle and the vast majority of blockchain projects cannot survive without oracles because it will directly make them blind.</p><p>As we all perceive, blockchain is a closed black box with no capability to initiate network call, but smart contracts are required to validate the data by third-party according to the consensus mechanism. To use a simple analogy, <strong>the role oracle played is the ambassador between two civilizations, like “Sopho” between the humans and the three bodies in “The Three-body Problem ”.</strong> Only by the data feed of oracle can smart contracts access the deterministic information from the Internet and real world, including the stock price, exchange rate, final results of presidential election, etc.</p><h2 id="h-family-tree" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>Family tree</strong></h2><p>In terms of format, oracles can be classified as software oracle, hardware oracle, centralized oracle and decentralized oracle:</p><h2 id="h-1software-oracle" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>1.Software Oracle</strong></h2><p>Software oracle is connected to the Internet and can access, transmit and input data and information from any third-party servers or websites by API, such as commodity prices, weather index, flights number, into smart contracts in real-time.</p><h2 id="h-2hardware-oracle" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>2.Hardware Oracle</strong></h2><p>Hardware oracle is widely adopted in IoT, working as the electronic sensor and data collector. It transfers physical events into digital values so that smart contracts can understand them. Barcode scanners, bank card POS machines, medical equipment for collecting all kinds of medical data, etc can all be identified as hardware oracles.</p><h2 id="h-3centralized-oracle" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>3.Centralized Oracle</strong></h2><p>Centralized oracle is a type of oracle with single-source data, usually backed by trustworthy third-party like government or reputable corporations. It can make data tamper-proof and loss-proof by separating data from untrusted operating systems of local equipment. Yet the sole source of data also brings certain vulnerabilities towards smart contracts.</p><h2 id="h-4decentralized-oracle" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>4.Decentralized Oracle</strong></h2><p>Decentralized oracle refers to an oracle with a distributed consensus mechanism, also known as consensus oracle. It is more reliable and trustless since it achieves data from multiple external sources rather than a single source. Regarding to the data processing methods, decentralized oracles can be categorized into 4 sectors based on the theory of Huawei Lab:</p><ul><li><p><strong>Aggregation-based processing:</strong> It eliminates single malicious data impact by multi-data source aggregation, i.e. Chainlink</p></li><li><p><strong>Staking-based processing:</strong> It requires participants to hold assets for greater credibility, i.e. Band</p></li><li><p><strong>Game-theory-based processing</strong>: It provides non-adversarial economic incentives, i.e. NEST</p></li><li><p><strong>Reputation-based processing</strong>: It restricts adversarial nodes by reducing their reputations, i.e. Witnet</p></li></ul><p>Compared with centralized oracle, decentralized oracle is less likely to bring security risks since it defused the sing-node failure issue, though its operation efficiency is relevently lower. Most DeFi applications prefer running on decentralized oracle based on such concerns.</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/7f7b5f51b6c58737a29870aef9d69a1387159e57a0c4cbc8bc49d78493234c16.png" alt="Source: IOSG" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="">Source: IOSG</figcaption></figure><h2 id="h-oracle-in-web30-transcending-defi-infra" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>Oracle in web3.0: Transcending DeFi infra</strong></h2><p>Based on the cognition above, we have an essential understanding towards the significance of oracle adopted in blockchain. But when it comes to the application value of oracle in web3.0, we are somehow still viewing it through a blurred lens.</p><p>Last year, the concept of <strong>DON (Decentralized Oracle Network)</strong> was first introduced in the Chainlink2.0 whitepaper. DON is a network maintained by a group of Chainlink nodes and can bootstrap any oracle services flexibly, so that in the near future Chainlink would be able to provide external data to the blockchain via completely trustless off-chain computation. To realize this vision, Chainlink has launched a series of products and services such as VRF, Keepers, CCIP and <strong>the deployment of those services opened up our imagination to the application scenarios of oracle in web3 to a large extent.</strong></p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/f23cd94e3dc97852591f13367b6070d98a17a25dd07c49e156c4ac6765c8516e.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>Source: Chainlink 2.0 Whitepaper</p><p>To better demonstrate the immense potential of oracle in future web3 world, we briefly illustrate here with a few examples and discuss them based on different scenarios:</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/1f7f1481a999f3f933eee552384b53ed0e4427f34beaf2f930e8c989bb0c3032.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>Source: IOSG</p><h2 id="h-defi" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>Defi</strong></h2><p>Though nowadays native cryptocurrency and stable coins are the mainstream in DeFi, predictably a novel market based <strong>tokenized real-world assets</strong> will be more and more appealing. For example, foreigners without legal identity in the US in such scenarios are able to purchase tokens pegged to certain US stocks. Also, users can extract liquidity by mortgaging tokenalized properties.</p><p>Another emerging usecase of oracle in DeFi is the** fair sequencing**. In the existing transaction regime, transactions are censored and ordered by the miners, which leaves great arbitrage and manipulation space to make use of. By exploiting information about upcoming transactions submitted to the mempool but not yet on-chain, miners and validators can modify the transaction sequence to benefit themselves. The term MEV(miner extractable value) is consequently generated to describe such phenomena.</p><p>Hence, to sort such issues out, oracles like Chainlink has developed certain solution called Fair Sequencing Services (FSS). FSS is poised to help DEX achieve fair transactions by designing deterministic algorithms so as to prevent MEV, front-running or any other schemes causing transaction disorder. **It mainly consists of three sections: transaction monitoring, transaction sequencing and transaction posting. **You can get a quick peek at the FSS mechanism via the schematic below:</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/ea7d0ffa5e0713ad05bab53c3bbcd7fbb8289728cb2d260bcdfc1069ff56134b.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>Source: IOSG</p><h2 id="h-nft-and-gamfi" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>NFT &amp; Gamfi</strong></h2><p>The NFT issuing mechanism and gamefi user experiences can also be enhanced significantly by the adoption of oracle. A typical usecase is the <strong>generation of verifiable random numbers based on the off-chain external data</strong>. In one case, artists could generate NFTs with varying rarity based on a verifiable random number and ensure fairness in airdrops. In another, game vendors could also use random numbers to generate more diversified combat gear and battle scenarios.</p><p>Academically speaking, random numbers are usually recognized as statistically randomly generated numbers, playing an integral role in cyber security, games, scientific simulations and many other critical fields. The generator that produces random sequences is called Random Number Generator (RNG). According to the nature of the random sequences generated, it can be divided into two categories: <strong>True Random Number Generator (TRNG) and Pseudorandom Number Generator (PRNG).</strong></p><p>TRNG generally utilizes uncertain physical phenomena such as noise, chaos, and quantum stochastic processes to generate true random sequences. Whereas, PRNG is a definitive algorithm that requires an initial external input to act as seed. Algorithms including Linear congruential, Cryptography, ANSI X9.17, Mersenne twister, etc. are relatively popular ones.</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/406269d656fa2f5bf650847ecc9c515bddd2b562599b96dd20e811b17ca17d47.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>Source: IOSG</p><p>Apparently, the random number generated on-chain is the pseudorandom number. Also, **given the transparent nature of blockchain content, random numbers generated on-chain are prone to be attacked by dishonest nodes and thus face considerable security risks. **Typically, miners can gain comparative advantages in gambling by dropping unfavourable blocks through “Block Withholding Attacks”.</p><p>Apart from boosting the cost of miner attacks through repeated hashing, the most expedient solution is to** employ oracle as a TRNG to produce verifiable true random sequences by integrating off-chain data**, and so as to ensure the confidentiality and impartiality of random numbers. At current stage, except approaches like Randao’s Commit Reveal and BLS contract, the <strong>VRF(Verifiable Random Function)</strong> provided by decentralized oracles are the most common services, which can be simply demonstrated in the following workflow:</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/501ec403f23ece4cf918ef1f7cd1608f99e4a172b3a488a01d9573f11c6f3ef7.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>Source: IOSG</p><h2 id="h-socialfi-and-dao" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>SocialFi &amp; DAO</strong></h2><p>In some cases, oracle can also act as a <strong>Decentralized identity (DID) tool</strong> in application scenarios of socialfi and DAO. By summoning DON to assemble Internet and off-chain activity data, users are empowered by oracles to verify and manage their own identity credentials with legacy compatibility and privacy preservation. For example, by integrating off-chain information on the activities participation and accreditation of DAO members via oracle, DAO administrators are capable to issue the corresponding POAP (proof of attendance protocol) and certify the competence qualifications of members.</p><p>And CanDID is the inside tool that helps DON to realize such functions. <strong>By adopting an oracle, CanDID allows users to import their identities from existing systems securely and prevents the possible duplication of identities.</strong> For instance, Alice and Bob can use their Social Security Profile pages to generate a certificate attesting to their Social Security numbers. In terms of mechanism, CanDID is mainly composed of two subsystems: <strong>identity system</strong> and <strong>key-recovery system</strong>.</p><p>In the identity system, CanDID can enable secure identity migration across existing web services such as social media, e-banking accounts, etc. by leveraging oracles especially like DECO or Town Crier. Additionally, CanDID doesn’t require data vendors to explicitly create DID-compatible credentials, which tremendously facilitates the convenience of the credential ecosystem.</p><p>In the key-recovery system, CanDID allows users to exploit existing network authentication schemes to recover keys through agile and simple workflows. Users can store keys on whatever device they use regularly and pre-select recovery policies and pre-store them via secret-sharing. The brief recovery process is shown in the following figure:</p><figure float="none" data-type="figure" class="img-center" style="max-width: null;"><img src="https://storage.googleapis.com/papyrus_images/8fea972cfcf23016a5ca77f60b3b519c323233c008aa5046b2391d44fa83e42a.png" alt="" blurdataurl="data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACwAAAAAAQABAAACAkQBADs=" nextheight="600" nextwidth="800" class="image-node embed"><figcaption HTMLAttributes="[object Object]" class="hide-figcaption"></figcaption></figure><p>Source: CanDID: Can-Do Decentralized Identity with Legacy Compatibility, Sybil-Resistance, and Accountability</p><h2 id="h-looking-forward" class="text-3xl font-header !mt-8 !mb-4 first:!mt-0 first:!mb-0"><strong>Looking forward</strong></h2><p>To conclude, it’s well perceived that oracle has opened a new chapter in the web3 era, ready to go beyond its restricted envision and obsolete functions. Whilst discussion on oracle’s burgeoning applications in web3 is still sparse, we expect that as the technology continues to evolve and crypto gradually engulfs the world, more and more savvy players will acknowledge the potential value of oracle and its significant contribution to web3.</p><p>Discover the undiscovered. Imagine the unimagined. More usecases and scenarios like data feeding innovations, and cross-chain solutions will be elaborated in our subsequent research.</p><p>(Acknowledgements: Chainlink China)</p>]]></content:encoded>
            <author>sallygu@newsletter.paragraph.com (sallygu.eth)</author>
        </item>
    </channel>
</rss>