本周摘要
$2 trillion的意义
Gitlab上市
NFT的发展
DeFi 2.0
Modular Blockchain
其他
$2 Trillion的意义
Crypto的资产在过去5年从$16 billion发展到现在的$2.3 trillion。Bank of England认为这会是一个潜在的系统风险,因为2008年的时候,次级贷款也就是$1.2 trillion。奇怪的是,为什么不在crypto市值超过$1.2 trillion的时候出来讲这个问题。
另外一个有意思的数据是,目前在private market里面,有800个unicorn,市值超过$2.6 trillion。和crypto体量接近,怪不得都是crypto is the new VC。另外,这800个独角兽全部上市,需要多久?
Gitlab上市
GitLab周四上市,首日上涨35%,市值接近$15 billion。
2018年,微软用$7.5 billion的价格收购Github。当时大家都觉得很贵。
Github的原高管有很多感慨。
Tesla的AI负责人Andrej Karpathy曾经写过一篇文章,认为Github代表的是Software 1.0,而ML是Software 2.0。
所以2.0是需要新的DevOps工具。目前这个领域叫做MLOps。里面的一家创业公司Weights & Biases最近估值超过$1 billion。
According to Weights & Biases co-founder Lukas Biewald, the software world has a set of tools built for developers to write and deploy code well. This could include a git-style service (GitLab, GitHub, etc.), monitoring (Atlassian, Datadog, etc.), and the like.
His company’s goal, he explained, is to build a similar stack of services for the ML world. And today, he explained, many ML teams are working with ad-hoc tooling or simply without software assistance.
The need for such a stack could be stark. A difference between the development and the ML world, according to Biewald, is that while code crashes when it fails, ML work can “behave badly” in more subtle ways.
除了工具,AI的training成本也是一个大问题。像GPT-3这样的模型培训成本超过500万美元,不是小团队能够负担的。所以需要在硬件和软件上想办法。
At the same time, the rising cost is pushing people to look for more efficient ways of training AI algorithms. Dozens of companies are working on specialized computer chips for both training and running AI programs.
Qi of Glean and McCreary of Optum are both talking to Mosaic ML, a startup spun out of MIT that is developing software tricks designed to increase the efficiency of machine-learning training.
NFT的发展
Visa发展NFT的incubation。
Visa is launching a non-fungible token (NFT) support program to help artists join the digital art space.
The financial services giant is partnering with former Major League Baseball player turned NFT artist Micah Johnson to build a program that will support artists who want to use NFTs to sell their work, Visa said in a press release.
FTX开始支持基于Solana的NFT项目。
Coinbase也准备开启NFT的销售,1天内有超过1百万的waiting list注册。目前OpenSea的累积交易用户是500k左右。所以Coinbase这个ramp还是不错的。
2.0的代表自然要走在3.0的对立面。只有2.0的受害者会积极拥抱
Valve bans blockchain games and NFTs on Steam, Epic will try to make it work
既然NFT这么火,那么自然会被讨论是否属于监管范围。
However, if you’re fractionalizing, and subsequently borrowing or lending against an asset and generating interest, you’ve created a financial product. Accordingly, these products should be registered as securities and subject to the same rules as if they were in the real world.
DeFi 2.0
DeFi 1.0的时候,主要通过liquidity mining来吸引外部资金。但是一旦资金撤走,项目就很容易垮掉。所以现在DeFi 2.0有了新的玩法,比较有代表性的是OlympusDAO。
第一步是通过Liquidity Pair来增加买量,比起纯的liquidity,LP要求资金方一定会购买token。
At any time, Olympus is willing to pay you OHM tokens in return for selling them liquidity. In this case, you get a 1.52% return over 5 days for selling them your OHM-FRAX LP. So instead of getting $5,000 of OHM if you had bought it directly, you’ll now get $5,076 worth of OHM! Sure you’d earn some fees holding the LP tokens, but you won’t make 1.52% in 5 days.
第二步是stake token来减少卖量
New OHM tokens are constantly being minted to pay people for buying liquidity bonds. If you just hold your OHM tokens, they should be getting less valuable over time as more and more are minted.
尽管token的总量不断在增加,但是如果token的价格是上升的,那么就是很强的feedback loop。那么,反之亦然。目前OHM的价格是这样
有意思的是OlympusDAO对自己的定位是decentralized reserve currency。
Modular Blockchain
由于DeFi和NFT的大热,ETH的transaction fee被认为是一个痛点。也就有了Polygon,Solana,Avalanche这样的项目。
但是之前的Solana的宕机事件就可以发现Solana未必是一个长期的解决方案。
如果blockchain需要被很多人使用,需要的是从monolithic的单链结构变为modular的结构,从vertical scaling变为horizontal scaling,
基本的共识是以roll-up为主,短期是Optimistic,长期是Zero Knowledge。
如果是按照这个框架来看,似乎ETH以为的L1空间对会有限,而L2的价值会更偏向于utility,而不是money。
Offchain Lab目前已经在考虑下一个版本的Arbitrum了。
How much faster? We’ll need to test it at scale to get a firm benchmark on the performance gain, but we’re estimating a 20–50x increase in Layer 2 execution speed, and we also expect a sizable decrease in costs.
有意思的是准备把结构改为Web Assembly架构
One of the decisions that we made early on when building Arbitrum was the focus of EVM compatibility and replicating the Ethereum developer experience. Since Arbitrum’s academic origins in 2013, we pioneered interactive fraud proofs, and our design gave us extreme flexibility in designing a Layer 2 platform that was fully compatible with the EVM. About 14 months ago, we began an exploratory research project asking if we could build a version of Arbitrum powered by the popular WebAssembly (WASM) architecture. The potential benefits were clear: a version of Arbitrum that was much faster and also even more compatible with the EVM — because existing EVM engines can run on WASM.
在roll-up为主的思路下,那么cross-chain会很有价值,dAMM就是一个应用。
现在有一个类似的项目,0x Protocol
其他
Grocery模式的比较。3P很难赚钱,1P改变了价值链。