Zack6.eth

Web3 Explorer

[Zack6.eth @0xzack6]

🚀 Thanks for supporting @TrustaLabs !

If you have been in Trusta-related campaigns across the ecosystem, please link your Starknet & TON accounts to the EVM address.

⏰ Link before [Jul 19 2025 09:00 UTC] & Prepare for the magic!

#Trusta

👉 https://t.co/SNfd4yTKcR

[ginobefun @hongming731]

Hacker News 上《我认识的最优秀的程序员》的文章，写得很实在，挺有共鸣。作者总结了他见过的那些顶尖程序员的共同点：

1. 死磕官方文档 (Read the Reference)： 这点我觉得太对了！遇到问题，别老是第一个就跑去 Stack Overflow 或者问 AI，也别瞎猜。大佬们会直接去看官方文档、去看源码。那才是第一手资料，而且很多时候写得比想象中清楚。

2. 深入理解工具 (Know Your Tools Really Well)： 光会用不够，得真正懂你用的技术。不光是知道怎么操作，还要了解它的历史、原理、局限性、生态。这样才能用得好，配置起来也心里有数。

3. 认真读报错信息 (Read The Error Message)： 别扫一眼就过！大佬们会仔细看报错，琢磨里面每个词的意思。作者说，很多时候答案就在报错信息里，看懂了就能解决大部分问题。

4. 拆解问题 (Break Down Problems)： 遇到难题卡壳了怎么办？高手会把大问题拆成一堆小问题，直到每个小问题都好解决。这基本就是程序员的核心工作：拆解。

5. 不怕“脏活累活” (Don’t Be Afraid To Get Your Hands Dirty)： 顶尖的开发者不怕钻研陌生的代码库，不会说“这个我不熟”就不管了。他们就是上手去搞，边搞边学，不知不觉就成了这块的专家。

6. 持续学习，不瞎跟风 (Never Stop Learning)： 技术更新快，大佬们会一直学习新东西，但不是盲目追潮流。他们会认真评估新技术的优劣，如果不用，也能清楚说出为啥，以及替代方案是啥。很多大佬年纪不小了，但思维还是很活跃。

7. 不耻下问，也不怕说“我不知道” (Don’t Be Afraid to Say “I Don’t Know”): 承认不知道不是丢人的事，反而是学习的开始。大佬们不怕暴露自己的知识边界，而且能坦诚地跟人交流，哪怕对方是刚入行的新人。

8. 拒绝猜测 (Don’t Guess)： 遇到不确定的地方，别凭感觉猜。去查证、去问、去调试。猜对了可能让你养成坏习惯，猜错了直接导致 bug。

9. 保持简单 (Keep It Simple)： 高手写的代码往往是简单、清晰、易维护的，而不是搞一堆花里胡哨但没人看得懂的“炫技”代码。

总的来说，强调基本功要扎实、持续学习、保持好奇心、动手能力强、沟通坦诚、解决问题思路清晰。感觉这些点都挺实在的，不是什么虚头巴脑的东西，分享给你瞅瞅！

[Trusta AI @TrustaLabs]

Yes, so do I!
Tips for Trusta identity agent show up in April🤔 @cz_binance

[Alexon @0xAlexon]

Infini：支付的皮囊，资管的心
Infini: A payments shell, An asset management core.

最近找我问支付的人多了起来，实在不想反复讲同一套的东西，还有新的事情等着我们做。又恰逢infini 被盗50m usd，看到他们全额赔付（虽然是应该的），想聊表心意支持，于是借花献佛，把聊支付的事情用infini举例来聊。

恰巧23年底24年年中的那阵子一直在琢磨支付这事，想做产品，因此有点观察的心得。如今支付和火热基本原因都是一样的：

1）收入上，Crypto领域能产生收入的东西有限 。
2）宏观上，美国确立了Crypto进入合规资产的宏观面改善。

Coincidentally, towards the end of 2023 and into mid-2024, I had been pondering the payment space and considering building a product in this area, which gave me some insights. The reasons behind the current buzz around payments are largely the same:

-The crypto space has limited avenues for generating revenue.

-The U.S. has established a macro-level improvement in crypto’s status as a compliant asset.

首先在我们聊支付这个事情的时候，我们首先明确我们聊的是什么。目前大部分我们讨论的加密支付都处于【收款产品】，看起来似乎【收单产品】在加密世界里目前并不受待见，我记得Sphere貌似是做收单的。

也非常好理解，收单产品确实是一门拼流量拼渠道的生意，门槛不高，传统支付里面能玩出花来也就是在多级分销，比如POS机信用卡的各种花活；美团通过小餐馆渗入收单转移成本给广告费等等。加密收单目前前端收费一般也就是0.02%-0.05%，扩张起来目前不是特别标准。

收款产品自然而然就变成了加密支付创业者的第一站，也是我们今天聊的重点。需要分成To B & To C。

First, when we talk about payments, we need to clarify what exactly we’re discussing. Currently, most of the crypto payment discussions revolve around ‘Receiving products’. It seems that ‘Acquiring products ’(merchant-side payment solutions) are not particularly favored in the crypto world. I recall Sphere is one of the few working on acquiring products.

This is quite understandable. Acquiring products are essentially a business that relies heavily on traffic and channels, with relatively low barriers to entry. In traditional payments, the only real innovation lies in multi-level distribution, such as the various tricks with POS machines and credit cards, or how platforms like Meituan penetrate small restaurants to shift costs to advertising fees, etc. In crypto, the front-end fees for acquiring products are generally around 0.02%-0.05%, and the expansion process isn’t particularly standardized yet.

As a result, Receiving products naturally become the first stop for crypto payment entrepreneurs and the focus of our discussion today. These can be divided into To B and To C segments.

To B端：

多数目前在盈利的公司都集中在to b的领域，他们更像是web2.5的版本。以Epay为代表的To B结算换汇，水上部分的客户多来自云服务、全球化的MCN等等，水下部分你就自己琢磨什么业务需要全球收款支付呗。

这个部分略有擦边，但是经营主体隔绝的好，实际KYC过程里只要合规，真有点什么事能够提交完整全套资料，合规方面压力还是可以接受的。

最近被拉出来讨论的 @Sidekick_Labs 这种商业模式也是非常典型的 To B收款公司会对接的客户。

总体大白话的翻译，就是收各种货币最后同意给你结算成USD，这是传统支付。这些Web2.5产品不过加了一道 USD-USDT，实际的业务场景就发生了非常大的转换。

基本的利润来源也都是USD-USDT的交易对。（毕竟传统法币的转换已经卷的毛利低的像第一产业）

估计infini后期也得做这块业务，毕竟这是目前唯数不多在盈利的板块，且切中垂直赛道就能具备盈利能力。比如有抢占广告投放代理业务的（大量的擦边、色情、当地合规的博彩牌照是收入端是USDT，但是投放费用是USD）。

调研过程中更有意思的是，发现了非常多Facebook Google的广告代理商手上有大量的USDT储备（因为通常他们还会垫款，垫资能力和返点都是一个大广告代理商的核心竞争力）。这些广告代理商手上的USDT很多都可以做小基金了。

有关 To B的问题推特上可以去找 @peach123as ，很多加密项目的OTC发工资是他做的，之前是某跨境支付的销冠。

Most companies currently turning a profit are concentrated in the To B space, operating more like a Web2.5 version.

Take Epay, for example, which specializes in To B settlement and currency exchange. Their above-board clients often come from cloud services, global MCNs (Multi-Channel Networks), and the like.

As for the "underwater" clients—well, you can probably guess what kinds of businesses need global payment collection. This segment skirts the edges a bit, but as long as the operating entities are well-insulated and the KYC process is compliant, with all necessary documentation ready in case of scrutiny, the compliance pressure remains manageable.

A recent example that’s been brought up for discussion is Sidekick, a business model that’s very typical of the clients To B payment collection companies would serve.

In plain terms, these companies collect various currencies and eventually settle them into USD—that’s traditional payments. These Web2.5 products add an extra layer: USD to USDT.

This shift fundamentally changes the business dynamics, with the primary profit source being the USD-USDT trading pair. (After all, traditional fiat currency conversion is already so competitive that profit margins are as thin as those in primary industries.)

It’s likely that Infini will eventually venture into this space, as it’s one of the few profitable sectors right now. By targeting a vertical niche, profitability becomes achievable.

For instance, some companies have seized opportunities in ad placement agency businesses (think large-scale gray-area operations, adult content, or locally compliant gambling licenses where revenue comes in USDT but ad spending is in USD).

What’s particularly interesting from my research is the discovery that many Facebook and Google ad agencies hold significant USDT reserves (since they often provide upfront funding—advance payment capabilities and rebates are core competitive advantages for large ad agencies). Some of these agencies hold enough USDT to start small funds.

For more on To B-related questions, you can reach out to @peach123as on Twitter. He’s handled OTC payrolls for many projects and was previously a top sales performer at a cross-border payment company.

合规上，成本最低的结构就是香港主体+加拿大主体分别取得 MSO牌照和MSB牌照。

MSO目前处于略微通缩的状态，每个月发放10张左右，每个月销毁的比发放的略多个一两张。申请费用不高，但是流程复杂，对于加密公司来说不如直接找中介。

当时我们咨询的中介如今在给Infini办理MSO，现在也时常找我们聊一下加密支付，毕竟还是有点差异。虽然买卖没成，但是交了个朋友，这事让我觉着世界是真小，当时他们还非常拒绝接触加密的客户。

当然也可以直接买，价格非标，但是海关的年审比较麻烦，尽量能申请就申请把。MSB牌照由于比较便宜和简单，如果这个都搞不定，就不用做这个业务了。

Infini这个项目不是不想做To B业务，而是MSO申请下来的时间一般顺利也需要10个月。如果下半年OTC牌照颁布，估计 Infini 还需要继续投入香港OTC牌照。

To B业务的合规成本是硬成本，无法下降。 如果希望提高毛利做的更精细化，去长尾小国家做该项业务目前看起来是个这一两年的趋势，但是对全球化运营的能力要求更高，好处是账很好算清楚。土耳其、埃及、越南都是非常好的在增长的市场。

From a compliance perspective, the most cost-effective structure is to establish entities in Hong Kong and Canada, obtaining an MSO (Money Service Operator) license and an MSB (Money Services Business) license, respectively.

The MSO license market is currently in a slight deflationary state, with around 10 licenses issued each month, and slightly more being revoked. The application fees aren’t high, but the process is complex. For crypto companies, it’s often more practical to go through an intermediary.

The intermediary we consulted back then is now handling MSO applications for Infini and occasionally reaches out to discuss crypto payments, as there are still some nuances. Although we didn’t end up doing business together, we became friends, which made me realize how small the world really is. At the time, they were quite hesitant to engage with crypto clients.

Alternatively, you can buy an existing license, but prices are non-standard, and the annual customs review can be quite cumbersome. If possible, it’s better to apply for one.

As for the MSB license, it’s relatively cheap and straightforward. If you can’t manage to secure this one, you probably shouldn’t be in this business.

So, it’s not that Infini doesn’t want to pursue To B business—it’s just that the MSO application process typically takes around 10 months even under smooth circumstances. If the OTC license regulations are rolled out in the second half of the year, Infini will likely need to invest further in securing a Hong Kong OTC license. Compliance costs for To B businesses are fixed and unavoidable.

If the goal is to improve gross margins and operate more efficiently, expanding into smaller, long-tail countries seems to be a trend for the next year or two. However, this requires stronger global operational capabilities. The upside is that the financials are easier to calculate, and markets like Turkey, Egypt, and Vietnam are showing promising growth.

To C部分：

目前都集中在信用卡这个领域，也是目前门槛相对较低的领域。此处我形容的门槛相对低指的是找一级发卡商合作，Infini目前应该也是用这个方式找了Mastercard的某一级发卡商。

卡组织的一级发卡商门槛是非常高的，如果我没记错Visa的一级应当只有十几家？

这个领域被诟病利润低的核心原因是卡组织吃掉了绝大部分利润，去看看Visa股票走势把，点击月线，您会回来评论。

这中间用户支付的大约2%的手续费需要分给卡组织、一级（通道）、二级（实际发卡），一年多以前的成本大概是1.8%左右，卡组织吃掉大部分。如今开始普及之后看到了大量1%左右手续费的信用卡，我听说是降低下来了，Visa也批了很多金融创新的新发卡商收了较低的费用。

我一直觉得加密支付是应该去掀卡组织的桌子，而不是寄人篱下。（但是这个阶段冷启动用这个手段获取客户是没有问题的）

也有很多其他创业者在绕开卡组织高额费用，例如 @_MapoTofu_ 在巴西和泰国直接和银行合作发的卡是不需要过卡组织的，本质都是找一个背书承兑方，适合地域性扩张，在当地有成本优势。

这个领域里面本质是抢占公域流量转化到私域流量，用谁都一样的情况下，不是选bybit有返现，就是选社区更亲近的infini。

To C :

Currently, the focus in the To C space is primarily on the credit card sector, which also happens to be the area with relatively low barriers to entry. When I say "low barriers," I’m referring to partnering with primary card issuers. Infini, for instance, seems to have taken this route by collaborating with a primary issuer for Mastercard.

The threshold to become a primary issuer for card networks is extremely high. If I recall correctly, Visa has only around a dozen primary issuers globally.

The core reason this sector is criticized for low profitability is that card networks take the lion’s share of the profits. Just take a look at Visa’s stock performance—click on the monthly chart, and you’ll come back to comment on this.

Out of the roughly 2% transaction fee paid by users, the card network, primary issuer (gateway), and secondary issuer (actual card issuer) each take a cut. A little over a year ago, the total cost was around 1.8%, with the card network swallowing the majority. Now, with wider adoption, we’re seeing credit cards with fees as low as 1%. I’ve heard that costs have come down, and Visa has approved many new fintech issuers at lower rates.

I’ve always believed that crypto payments should aim to disrupt the card networks rather than rely on them. However, at this stage, using this method to acquire customers during the cold-start phase is perfectly reasonable.

There are also other entrepreneurs finding ways to bypass the high fees charged by card networks. For example, @MapoTofu has partnered directly with banks in Brazil and Thailand to issue cards that don’t go through card networks. The essence here is finding a trusted endorser or acceptor, which is suitable for regional expansion and offers cost advantages locally.

At its core, this field is about capturing public traffic and converting it into private traffic. When it comes to choosing a service, it often boils down to whether users prefer Bybit for cashback or Infini for its closer community ties.

单一业务来说，确实这个业务属于风险高于收益的项目。（比如 infini 不赚钱交朋友，莫名其妙亏了本金；或者让热心网友 @ohyishi 来说一下 onekey 的卡为啥停了的二三故事）

但显然infini的思考原点不是我要做信用卡，他们的思考原点跟我们当时是接近的，在我们一直觉得基金这个模式有问题的时候，我们一直琢磨什么东西能：
1）用户会用
2）能吸储做资管。

思考过后确实认为支付具备这个可能，只是我们在思考我们优势在做资管而不是直接竞争支付，前期的牌照过程太长，OTC牌照在当时还是个影子。

当然，也怕前端有 @Christianeth 这种选手杀出来（确实也发生了），技术门槛不高又需要网络效应的事情最怕有钱的铁子过来直接平推。

所以，从他们的招聘上来看，应当思考也是一样的，最后都要走向做资管的路线，但是由于没有MSO，开不了VA账户目前，所以妥协的从社区开始做起，做个基础增长再来聊以后。所以，你如果把infini当作一个支付发卡的公司，我认为它几乎没有资产价值，毕竟是二级卡商。

From a single-business perspective, this venture indeed falls into the category of high-risk, low-reward projects. For instance, Infini might not be making money but is building community, or it could inexplicably lose its money. Or, as the enthusiastic netizen @ohyishi might share, there are a couple of stories about why OneKey’s card service was discontinued.

However, it’s clear that Infini’s starting point wasn’t simply to issue credit cards. Their initial thinking was likely similar to ours back then. While we were questioning the sustainability of the fund model, we were also pondering what could ：

1) be widely adopted by users
2) serve as a vehicle for attracting deposits and managing assets.

After some thought, we concluded that payments could indeed fulfill this role. However, we recognized that our strength lay in asset management rather than directly competing in the payment space.

The licensing process was too lengthy, and the OTC license was still a distant possibility at the time. We were also wary of competitors like Chris emerging (which indeed happened).

In areas with low technical barriers but high network effects, the biggest threat is a well-funded player coming in and dominating the market.

Judging from their hiring strategy, it seems Infini is thinking along the same lines—ultimately aiming to move into asset management.

However, without an MSO license, they currently can’t open VA accounts, so they’ve had to compromise by starting with the community, building a foundation for growth before discussing future plans. If you view Infini purely as a card-issuing company, I’d argue it holds little asset value, given that it’s essentially a secondary card issuer.

但是如果你的视角是资管公司包了个支付的壳，那思考的角度是要变化的。关键的问题就是变成了以下三个：

1）作为支付的流量入口怎么在尽可能低的成本下跑增长 2）作为资管的实际收入口径如何稳定、安全、持续的赚钱 3）管理规模能有多大

写了那么多，这三个问题才是正确infini的发问方式。

3）最吸引人，目前稳定币市场持续扩张是显学，我看有人说这是个千亿美金容量，也没错，毕竟整个香港OTC都这么说，包括Hashkey和OSL。

题外话，Hashkey和OSL目前最大的营收也是来自于OTC，属于拿了VASP牌照但由于政策原因做成了USD-USDT OTC的生意，维护成本比MSO+MSB贵数个数量级。

1）问题怕是需要他们另一位合伙人 @0xsexybanana 来回答
2）问题看起来是需要招聘或者合作

所以别再讨论信用卡业务了，不过是明修栈道，暗度陈仓。但是2可能是最难回答的问题，所以我推测的思考路径是infini想和BSC站在一起，原因是Defi管理大量资金的黑天鹅是有概率的，希望Cedefi把钱合作到Binance的机构账户里去一部分。

However, if you view it as an asset management company wrapped in a payments shell, the perspective shifts entirely. The key questions then become the following three:

As a payment traffic gateway, how can growth be achieved at the lowest possible cost?

As an asset management entity, how can revenue streams be stable, secure, and sustainable?

What scale of assets under management (AUM) can be achieved?

After all that’s been written, these three questions are the right way to frame Infini’s strategy.

Point 3) is the most intriguing. The stablecoin market is undeniably expanding, and some estimate its capacity to be in the hundreds of billions of dollars. This isn’t far off, given that even Hong Kong’s OTC market, including Hashkey and OSL, echoes this sentiment.

As an aside, Hashkey and OSL’s largest revenue stream currently comes from OTC trading. Despite holding VASP licenses, regulatory constraints have turned their business into a USD-USDT OTC operation, with maintenance costs several orders of magnitude higher than those of MSO + MSB licenses.

Point 1) is likely a question best answered by Infini’s leadership.

Point 2) seems to hinge on either hiring the right talent or forming strategic partnerships.

So, let’s stop fixating on the credit card business—it’s merely a diversion, a feint to mask the real strategy.

However, Point 2) might be the toughest to address. My speculation is that Infini aims to align closely with BSC (Binance Smart Chain). The reasoning is that while DeFi managing large sums of money carries inherent black swan risks, the hope is that CeDeFi could channel a portion of funds into Binance’s institutional accounts for added security.

Cheems，部分原因应当也是买的Binance示好的可能性。

在BSC需要一个社区Meme的时候先交朋友后做生意，这是一个聪明的做法，往往喜欢上来算的很清楚谈判方式多数都是存量的生意或者小生意，需要增长的赛道更聪明的就是先付出，多发心，少问回报，相信对方是聪明人。

我和对方团队完全不熟，只有我加过郡主问Megaeth的NFT，想要一点。还有个叫Nathan的哥们加过我来交了个朋友，利益无关。

单纯觉着，被盗钱的影响肯定很大，希望他们渡过过去，重新审视业务安全，全额赔付这么大笔钱可能会丢了钱，但是赢了尊重。我们也赔付过，虽然不是这么大额。

坦率的说，赔付只是一时的，但后续长时间都要因为赔付会难免动作变形。个中苦楚，都是表面看不到的。

Cheems, part of the reason might also be to signal goodwill toward Binance. When BSC needed a community meme, they chose to build friendships first and do business later.

This is a smart approach. Often, those who prefer to negotiate with everything calculated upfront are dealing with存量生意 (static markets) or small-scale ventures. In growth-driven sectors, the wiser move is to give first, invest more heart, ask less about returns, and trust that the other party is equally intelligent.

I’m not at all familiar with their team. The only interaction I’ve had was adding @0xsexybanana to ask if there were any MegaETH NFTs left, as I wanted a few. There was also a guy named Nathan who added me to connect and become friends—no vested interests involved.

I simply feel that the impact of the hack must be significant, and I hope they can get through this, re-examine their security measures, and move forward. While reimbursing such a large sum might mean losing money, it wins respect. We’ve also reimbursed before, though not on such a large scale.

Frankly, reimbursement is just the immediate step, but the long-term aftermath often leads to operational distortions. The hidden struggles are never visible on the surface.

支付战争里面描绘了Musk和Peter Thiel的故事，可谁说这支付战争2.0里面就不会出日后书里的人物呢，是不过是谁来当乱世英雄罢了。

The Payments War depicted the stories of Musk and Peter Thiel, but who’s to say that in Payments War 2.0, future figures won’t emerge? It’s just a matter of who will rise as the hero in these turbulent times.

[Zack6.eth @0xzack6]

Yap early, yap only, yap often.

@KaitoAI is connecting AI, attention and capital with Yaps.

Just claimed my social card and I'm accumulating Yap points in real-time.

Claim yours 👉 https://t.co/HIrESUTlss

[Zack6.eth @0xzack6]

🚀 Just snagged my Pioneer Souvenir Cases from @Matr1xOfficial.

Check if you've got yours at https://t.co/1ijLvJFf2V

🎁 Limited edition exclusive to early contributors. And the unboxing—oh, it’s big - extra $MAX airdrops for the lucky ones!

#Matr1xCase $FIRE @Matr1x_io

[Linea.eth @LineaBuild]

https://t.co/Luej4kU0ur

[Deagent Network @DeagentNetwork]

Introducing: The New Era of Delegation

https://t.co/7dbhkWknwF

[MATR1X @Matr1x_io]

Emerging from the depths, @Castileofficial joins the @Matr1x_io realm as our third epic adventure.

Together, we'll weave a tale of power and mystery. A new chapter begins.

Stay tuned for our joint release.

#Matr1xUniverse @Matr1xOfficial @Matr1xBattle

[Taiko.eth 🥁 @taikoxyz]

We're happy to announce our latest campaign on @Galxe. Taiko x Blazpay: Mint, Stake, and Play! ⚡️

Check out @BlazpayOfficial's innovative NFT entrypass system and explore new dashboard features. 🎮

🔗 Start your adventure: https://t.co/ViCpFTkze0

Check out what's in store 👇

[Zack6.eth @0xzack6]

https://t.co/q74SfGvcnu

[io.net @ionet]

That's a wrap on @LEAPandInnovate 2024!

The https://t.co/ZuybGWvjv9 team made a splash at the largest annual tech conference in Saudi Arabia this year, from kicking off a partnership with @TheSandboxGame to building an even larger base of suppliers and customers for the network.

Watch the highlights from the event below! We look forward to seeing you at our next event - Token2049 in Dubai!

[io.net @ionet]

That's a wrap on @LEAPandInnovate 2024!

The https://t.co/ZuybGWvjv9 team made a splash at the largest annual tech conference in Saudi Arabia this year, from kicking off a partnership with @TheSandboxGame to building an even larger base of suppliers and customers for the network.

Watch the highlights from the event below! We look forward to seeing you at our next event - Token2049 in Dubai!

[Zack6.eth @0xzack6]

Verifying my Twitter account for my #GalxeID gid:5JH2zX3VyyegTizmMG4qd @Galxe

https://t.co/Wqom80va2L

[Zack6.eth @0xzack6]

Sometimes it's good for the $TEA to burn. TEA tokens were just burned on [manually insert basescan sepolia link to burn event] #BurnTheTEA https://t.co/9FfTT4mHoO

[Trusta AI @TrustaLabs]

🔥Ohhhh my dear Trustas!

🥳I'm thrilled to share the first exciting news of March. @TrustaLabs is on the verge of unveiling the OG Dragon🐉, an NFT exclusively associated with Trusta. Positioned as an OG pass, this remarkable digital asset is slated for release around April.
Let’s delve into the essential details:

#⃣Basic
Number: 3999
WhitelistA: 1500(free)
WhitelistB: 1500(0.015eth)
Public Mint: 800(0.05eth)
Eco-Ops Reservation: 199

🦹‍♂️NFT Rights and Benefits
1.Mining eTrusta Points: Each day, you can mine 15 eTrusta points. ✨These points can be daily collected from the official website without incurring any gas fees.
2.OG Pass Access: As an OG pass holder, you’ll enjoy priority access to future potential Trusta and partner benefits.😯

👇Whitelist Distribution:
Our Stance: We reject behaviors that consume users, including DC Liver White & etc. To get whitelisted, actively participate in various communities and keep an eye out for our High MEDIA Score Lottery!☑️

🌈We also extend a warm welcome to all contributors to Trusta. Let us know about you by applying here:🙋‍♂️
https://t.co/tJWWiuq3bd
🙏Due to limited availability, only part will be directly contacted. However, those not get contacted can still get a chance obtaining the whitelist through various community raffles.💦

🩷&RT, Spread your love!🫶

[Trusta AI @TrustaLabs]

Unveiling Trusta: Answers to Your Top Questions🫡

1⃣Demystifying Project Airdrops
Projects normally analyze multi-chain/single-chain MEDIA data (e.g. @CelestiaOrg @MantaNetwork ) . The specific dimensions and gradients are strictly confidential to anyone other than the project party even to us.

We exclude project-defined bots (using synchronizers, fund pooling, etc.) No one-size-fits-all bots label exists, because every project has their own standards.

2⃣Unveiling Media's Role in Airdrops
As a third-party service, Trusta Labs provides MEDIA dimensions for analysis, which stands for monetary, engagement, diversity, identity, age. Projects can choose which indicators and gradients they use, tailored to their needs. It is more like a “Modular Airdrop Standards” based on MEDIA. Welcome to click on MEDIA Analysis for details: https://t.co/d7478zRzEH

3⃣The sybil label shown on official website before
The first cooperation was with @gitcoinpassport , this sybil label is only for the Ethereum main network. It has nothing to do with the airdrops with any other program.
-2 means that the on-chain data is undetectable due to lacking activities. -1 to 60 can be passed, the higher the score represents the greater the sybil risk on that chain. And it is not open for inquiry.

4⃣The more Odysseys you participate in, the more likely you are bots?
On the contrary, participating in Odysseys reduces misidentification of genuine users, as most participate in the same contracts. And we can exclude these contracts to improve the algorithmic accuracy.
Do notice that both those who received and missed Starknet airdrops have participated in Odysseys. Actually it is really hard meeting "bots" criteria for regular users.

5⃣High MEDIA Score not getting airdrops
MEDIA Score summarizes your chain interactions for easy overview.
High MEDIA score indicates deep interactions with the chain, but it is up to each program to choose from MEDIA dimensions on which indicator&gradients they are going to use for airdrops.
Please also double check if you satisfy the basic rules~

6⃣ Addressing "Farmer" Concerns:
Trusta Labs hereby solemnly promises that it will never participate in any farmer work. Trusta Labs is committed to building AI+Web3 on-chain reputation system. In addition, in the airdrop activities like Celestia , MANTA, ASMATCH, Starknet, Trusta Labs only provide MEDIA dimensional analysis and POH. How can we participate in advance?

7⃣Beyond the Basics:
🤔Trusta Labs has never considered individual with multi-accounts to be a sybil, even if the purpose of the interaction is to obtain airdrop rewards.
However, utilizing tools and scripts to interact in bulk and compete unequally with real users for airdrop resources is against the original intent of most project parties and hurts the interests of normal ecological users.

🧑‍🤝‍🧑We are committed to building a fair and impartial Web3 environment. We want to provide a better on-chain experience for users and will consistently improve the MEDIA algorithm.
We welcome all Trustas to monitor and provide feedback on the work!🤝

[supermao @buaaxhm]

一个开发者成熟的标志是不再醉心于技术多么高大上多么酷炫多么牛逼。而是花大量精力去研究用户真的关心什么喜欢什么，并且思考怎么样用自己掌握的技术去真正的给用户创造价值。

[宝玉 @dotey]

转译：一种全新的 RAG 方法让文档理解更上一层楼

RAG 技术已经迅速成为在实际应用中部署大型语言模型（LLMs）的首选方式。但大部分现有技术仍面临一个局限：它们只能从文档中检索到一小段连续的文本，这限制了对整个文档上下文全面理解的能力。

现有一种创新方法，名为 RAPTOR（递归抽象处理树形组织检索），它通过递归地嵌入、聚类并总结文本段落，从底层到顶层构建出一个分层次总结的树状结构。

在使用这个模型进行推理时，RAPTOR 能够从这棵树中提取信息，实现对长文档不同层次抽象信息的整合。

通过实验对比，我们发现，相较于传统的检索增强语言模型，RAPTOR 通过递归总结的检索方式在多个任务上都取得了显著的进步。

特别是在需要复杂多步推理的问答任务中，结合 RAPTOR 的检索功能和 GPT-4 的强大能力，我们在 QuALITY 基准测试上的表现提高了20个百分点，达到了新的最佳水平。

未来，我们期待出现更多类似 RAPTOR 这样的方法，进一步深化对文档的理解并优化 RAG 技术。RAPTOR 的优势在于它专注于改善检索部分，而不是像其他方法那样，通过添加噪声或引入网络搜索结果来调整上下文。

论文：https://t.co/u7FAGyeeqC

[Zack6.eth @0xzack6]

make sense

[董币哥 @NextBullReady]

行动力不足来自于精确性不足

[Zack6.eth @0xzack6]

转帖保存，使用 https://t.co/Hwetms3JLn 这个安装更科学

[Tw93 @HiTw93]

#工程师工具 最近一年用得最多的 Mac 垃圾清理工具居然是一个命令行 mac-cleanup-sh，通过 curl -fsSL https://t.co/BYD7qi4ru7 | bash -s update 进行安装，然后在 .zshrc 里面加上 alias clean="sudo /usr/local/bin/mac-cleanup"，直接命令行 clean 就可以很方便使用，很方便～