Lao bai · 7

Tokens, Models, and Bubbles: The Crypto × AI Game in the Tier 1 Market

Author: Lao Bai Original title: Looking at Crypto × AI: A Tokenization Illusion Experiment After a lapse of 2 years, V posted Twi again. I'll follow the research report from two years ago. Even the time was exactly the same, February 10. (Related reading: ABCDE: Sorting out AI+Crypto from the perspective of the primary market) 2 years ago, God V vaguely expressed that he was not very optimistic about the various Crypto Helps AI we were popular at the time. The popular troika in the industry at the time was computing power capitalization, data capitalization, and model capitalization. My research report from 2 years ago was mainly about some of the phenomena and questions that this troika observed in the primary market. From the perspective of God V, he is still more optimistic about AI Helps Crypto. A few examples he gave at the time were: AI as a player in a game; AI as a game interface; AI as a game rule; AI as a game goal; in the past two years, we actually tried a lot of things on Crypto Helps AI, but there were few results. Many tracks and projects were all - just sending a coin, there was no real commercial PMF; I called it the “tokenization illusion.” 1. Capitalization of computing power - Most are unable to provide commercial-grade SLAs, are unstable, and frequently disconnected. It can only handle simple small to medium model inference tasks. Most serve marginal markets, and revenue is not linked to tokens... 2. Data capitalization - High supply-side (retail) friction, low will, and high uncertainty. On the demand side (enterprise), what is needed is a structured, context-dependent, professional data provider with trust and legal responsibility. Web3 projects that are DAO entities are difficult to provide. 3. Model capitalization - The model is inherently a process asset that is not scarce, can be replicated, fine-tuned, and rapidly depreciated, rather than a final asset. Hugging Face itself is a collaboration and communication platform, more like GitHub for ML, rather than App Store for models, so the so-called “decentralized Hugging Face” tokenizing models basically ended in failure. Furthermore, we have tried various “verifiable deductions” in the past two years. This is also a typical story of finding a nail with a hammer. From zkml to OPML to Gaming Theory, etc., even EigenLayer turned his restaking narrative into verifiable AI. But it's basically similar to what happened at the Restaking Track - few AVS are willing to keep paying for additional verifiable security. Similarly, verifiable reasoning is basically verifying “something no one really needs to be verified”, and the demand-side threat model is extremely vague - who is being protected? There are far more AI output errors (model capability issues) than AI output being maliciously tampered with (counterfeit problem). Some time ago, everyone also saw various security incidents on OpenClaw and MoltBook. The real problem comes from: the wrong strategy design gives more permissions, no clear boundaries, and there are almost no imaginary nails where “the model is tampered with” and “the inference process is maliciously rewritten.” I posted this picture last year, and I don't know if Lao Tie remembers it. The ideas given by God V this time are clearly more mature than two years ago, also because of our progress in various directions such as privacy, X402, ERC8004, and forecasting the market. It can be seen that half of the four quadrants he divided this time belongs to AI Helps Crypto, and the other half belongs to Crypto Helps AI, and is no longer clearly biased towards the front two years ago. Top left and bottom left - Using Ethereum's decentralization and transparency to solve AI trust and economic collaboration issues 1. Using ZK, FHE and other technologies to ensure the privacy and verifiability of AI interactions (I don't know if the verifiability reasoning I mentioned earlier counts). 2. Ethereum as an economic layer for AI (infrastructure + prosperity): Enabling AI agents (Agents) to conduct transactions through Ethereum...

191d agoLuxurytracy

ABCDE presses the pause button. What is the current state of survival of Chinese VCs?

Source: ChainCatcher Author: Nian Qing Recently, ABCDE Co-Founder Du Jun (Du Jun) officially published an article announcing that ABCDE Capital has stopped investing in new projects and suspended the second phase of the fund raising plan, and announced the launch of a new incubator product, Vernal. ABCDE has launched 19 times in the past year and isn't particularly conservative. Despite having plenty of ammunition, it still can't resist the current boring crypto market. After the FUD VC revolution began, Crypto VC began pressing the pause button under hell-level difficulty. Some chose not to take action and only keep core personnel to maintain post-investment services, and some VCs began to seek transformation to second-level investments, projects, incubators, and crypto marketing. Compared with European and American VCs in the same period, Asian VCs seem to be more conservative. ABCDE investment and research partner Lao Bai asked the question a month ago — why while we are “out of fire” in Asia, Europe and the US are still “firing”; why are they daring to pull the trigger in this valuation? In response to this, he speculates that the reasons may include different exit cycles, differences in investment decisions, and values, Asian VCs are more concerned about yield, and European and American peers are idealistic and long-term. Furthermore, according to investor data from the RootData platform, in terms of the number of launches in the past year, most VCs in Chinese-speaking regions have adopted a more conservative investment strategy, and the number of shots taken was less than 10. Of course, there are also leading venture capitals such as Animoca Brands, OKX Ventures, HashKey Capital, SNZ Holding, Waterdrip Capital, Mask Network, Nomad Capital, and NGC Ventures, which still maintain high-frequency transactions, and have launched more than 20 times in the past year. Although IOSG Ventures, Fenbushi Capital, SevenX Ventures, LongHash Ventures, PAKA, etc. have launched more than 10 times in the past year, the number of investments has decreased significantly compared to the previous year. This article sorts out the current situation of major Chinese-language VCs/VCs founded by Chinese people that have had few launches in the past year. (Note: The data in this article is based on publicly disclosed investments. It is not ruled out that there may be undisclosed internal operations or investments, which may differ from actual investment data.) Continue Captial, founded by Continue Captial Coin Circle OG Pima (Pima), has invested only 0 in the past year. The most recent investment was in March of last year, and participated in io.net's Series A financing. Pima is a senior investor in the cryptocurrency industry. He entered the digital currency field in 2013 and has invested in many well-known blockchain projects. Currently, both the two co-founders and Continue Captial are in a “lying flat” state, and neither the official blog nor Twitter are operating. The latest news about Continue Captial was in October of last year. Lin Xiahong, the co-founder of Continue Capital, is suspected of having stolen 35 million dollars. Lin Xiahong himself wrote on the X platform in response: “I accidentally revealed that I haven't left the ring; I passively made charity headlines.” LD CapitalLD Capital has made only 8 investments in the past year. The most recent investment was in September of last year, which is a significant reduction in the frequency of previous investments. LD Capital also has businesses such as secondary market investments and trading funds. Its funds include Beco Fund, FoF, hedge funds, and Meta Fund. In 2023, LD Capital also established Cycle Ventures, an early-stage investment fund focused on Infra, AI, and Dapps. In response to the VC dilemma, Lee Lihua, founder of LD Capital, said that the decline of Web3 investors was due to changes in hedging rules, low investment success rates, and “good projects, undervaluation, and large amounts” constituting a “triangular problem” that investors cannot balance. He also pointed out that in the end, many investors only get a quota of a few points, and may not only bear financial losses, but also face reputational risks. Bixin Ventures, founded by OG Starry Sky in the Chinese currency industry, has invested only 5 transactions in the past year. The most recent investment was in January of this year, compared to 202...

487d agoWendy#ABCDE #Animoca Brands #HashKey Capital #Lao Bai #Mask Network #NGC Ventures #Nomad Capital #OKX Ventures #SNZ Holding #VC #Waterdrip Capital #Du Jun #venture capital

2024 Public Chain Infra Miscellaneous Discussion

This article summarizes that the most popular tracks in the Tier 1 market are AI and BTC, and the AI bubble is expected to peak next year. The author suggests paying attention to the changes that have occurred at the Infra level in recent months, especially the public chain Infra circuit. Among them, the most popular concept at the executive level is Parallel EVM, but different projects have different technical routes and ultimate goals. The author introduced several different technical routes of Parallel EVM, as well as the different genres that attach importance to Parallel EVM. On the DA circuit, several projects have begun to evolve towards the ZK verification/settlement layer, while there are also new developments in other areas such as Oracle and MEV and interoperability. Thanks @cyodyssey的审阅与修订. Author: Lao Bai Note: This article is from @Wuhuoqiu Twitter: Recently, in the primary market, the most popular circuit is undoubtedly AI, followed by BTC. 80% of the projects I talk about every day focus on these two tracks. Personally, I can talk about 5 or 6 AI projects a day at most. It is foreseeable that the AI bubble will reach its peak the next year. With the launch of hundreds of new AI projects, the market value of the AI circuit will reach its peak. In the end, the bubble bursts, and at the same time, there will be unicorns that actually find the right fit for AI X Crypto, and continue to push this track and the entire industry forward. So in the current environment where AI is overheated, calm down and see what changes have taken place at the Infra level in recent months, especially the public chain Infra circuit. ETH, or the further deconstruction of the single-chip chain, when Celestia first proposed the modular concept and DA layer concept back then, actually took quite a bit of time to digest and understand. Now the concept is already deeply entrenched in the hearts of the people. Various RaaS infrastructures have spread to an exaggerated stage of execution layer, DA layer, and settlement layer over the past few months. Each layer has spawned new technical solutions, and even the concept of the settlement layer is no longer unique to ETH. Let's pick representative technologies from each layer to briefly describe the following two. Execution layer - The most popular concept at the executive level is undoubtedly Parallel EVM - represented by Monad, Sei, MegaETH, etc., and some existing projects such as FTM and Canto have also begun planning to upgrade in this direction. However, just as not all ZK projects protect privacy, the projects labeled by Parallel EVM actually differ in terms of technical routes and ultimate goals. Take a picture of Sei as an intuitive illustration. It's obvious that in an optimistic situation, changing the current sequential processing to parallel processing improves performance or is very obvious. Parallel EVM can actually be divided into several different technical routes from the perspective of how transactions go parallel - nothing new under the sun, nothing but the difference between prior and posterior experience is represented by Solana and Sui, requiring transactions to clearly state which parts of the chain state they have modified, so that they can check whether there are any state conflicts before packaging the block (such as AMM access to the same AMM pool) ), in some cases, the post-test of discarding these conflicting transactions is also called optimistic parallelism, represented by Aptos BlocksTM, that is, assuming that no one is in conflict and then accepting the transaction, then testing it. If a conflicting transaction is found, the transaction is declared invalid. The result is refreshed, re-executed, and this step is repeated until all transactions in the block have been executed. Sei, Monad, MegaETH, and Canto use similar solutions. In the primary market, we have also seen parallelized solutions for state conflict situations (such as the same AMM pool access described above), but it seems that the project will be relatively complicated, uncertain whether it is commercially viable, and the degree of emphasis is placed on Parrounding EVM in the evaluation - it can also be divided into two genres, represented by Monad and Sei, as the main expansion idea, that is, parallelization is the main expansion idea Narrative, for example, in addition to optimistic parallel processing, Monad also has MonadDB, which specializes in parallel processing for asynchronous I/O. Another idea is Fantom, Solana, and MegaETH. Parallelization is one of the expansion plans, but it's only one. Parallelization is an auxiliary story...

861d agody zhang#AI #Celestia #Infra #Solana #Ethereum #Public chain #Bitcoin

Bitcoin's Tangled Tech Stack: History, Current Status, and Future

Author: Lao Bai Original Article Source: Twitter Originally, this post was supposed to talk about trends in derivatives, NFTs, and NFTFI, but given the popularity of the BTC ecosystem recently, I also recently looked at several primary market projects in the BTC ecosystem. Let's just talk about BTC's current technical stack and trends, which are full of controversy and controversy. 1. To understand the history of BRC20, Ordinal, etc., you must have a general understanding of their technical and historical background. I will introduce them here in the most plain language. At the beginning, Satoshi Nakamoto invented Bitcoin as a peer-to-peer cash system. Later, on the forum, he said on the forum that Bitcoin has been booming and growing, so it can carry more things on it, such as escrow transactions, bonded contracts, third-party arbitration, etc. However, later, everyone discovered that the performance of BTC was too poor, the price fluctuated greatly, and peer-to-peer cash was impossible, so it slowly moved towards today's “electronic gold” route. Satoshi Nakamoto's original intention and vision also laid the groundwork for later expansion disputes and the incident where BCH and BSV split. Bitcoin has an OP_return script operation code, which can insert 80 bytes of messy stuff, so someone thought of dyeing Colored Coin for 12 years and used Op_Return to mark something, so that Bitcoin could change from homogenization to non-homogenization, but later it really couldn't be done because of BTC's performance problems. Then, in '13, there was MasterCoin (later OmniLayer, BTC-based USDT was issued here), and in '14, CounterParty used this Op_Return to do things. As you can imagine, they are all scrapped (no one uses BTC for USDT now). One of the co-authors of dyed coins in '14, he made ETH himself out of anger when suggesting a BTC upgrade to MasterCoin was ignored. Since then, numerous disputes have arisen due to capacity expansion issues. The most famous ones are undoubtedly BCH and BSV, which both follow the big block route. BCH slowly grew bigger, from 1M to 8M, then 32M, etc. BSV was more extreme, directly up to several hundred M or even one G. On the BTC side, it is popular. It uses Segwit's segregated witness technology to separate signature verification from transactions. The original 1M total data is now 1M transaction data +3M witness data, disguised as a “small expansion”. Note, quarantine witness is important! Because of Ordinal, BRC20 all this stuff originated from here. 2. Problems and Current Situation In addition to direct block expansion, BTC's own technology stack also has many methods, but it's far less popular than ETH. After all, it supports smart contracts and doesn't support technical complexity and viability, but one problem does exist. That is, after 100 years, there will be no block rewards (there will be few after 10 years). What about the security of BTC, can it become an increase in POS? How to stimulate enough on-chain TX is a problem. Overall, the current exploration directions are as follows: 1. Pure sidechain - Liquid Network, this is an affiliate chain run by BlockSteam, that is, about 10 times the performance of BTC. It mainly targets institutions for large Bitcoin transfers and transaction settlements. 2. Status Channel - Lightning Network, needless to explain. Everyone in the industry knows that it is currently one of the most popular technology stacks for BTC. It supports small or even micropayment designs for quick payments. Walmart, McDonald's, etc. support it, but after a few years, the data is really average. It has locked around 5,000 BTC, or about 70,000 channels. 3. Disguised sidechains - RSK and Stacks. RSK is EVM compatible and uses a mechanism for joint mining with Bitcoin to generate blocks. Stacks are not EVM compatible. They created their own Clarity language and used a POX transmission to prove blocks. Both don't actually inherit the security or hashing power of BTC; they just look like it; it's completely unrelated to ETH L2. 4. Client-side verification - RGB and Taro are relatively new paradigms, combining off-chain asset issuance and on-chain UTXO...

1177d agody zhang#Atom #Bitcoin SV #BRC #USDT #Ethereum #Bitcoin #Bitcoin #Bitcoin cash

Hong Kong conference, rethinking 3 popular racetracks

Author: Lao Bai, ABCDE Original article source: medium The Hong Kong conference is in full swing recently, and the primary market is picking up. ABCDE Q1 watched more than 100 projects and experienced firsthand several particularly hot tracks in the market, the most popular of which include Appchain, ZK, and Gaming. Since this year, these 3 tracks have also undergone some changes, and we have also rethought and recorded it as a memo for everyone's reference. one. Appchain (especially RAAS, Rollup as a Service) RAAS is a circuit that started up at the end of last year, and has a lot to do with the release of OP Stack, but Appchain as a Service has existed for a long time, represented by the Cosmos SDK, and gradually became popular after Celestia proposed a modular blockchain. RAAS can be seen as one of the hottest subsets recently. Why is AppChain as a Service track likely in the near future It was overHype? First, if you are a developer and want to make an Appchain yourself, the options you face are as follows. If your chain is EVM, then you can do 1. Make a pure ETH sidechain like Ronin (I believe few people would do this now) 2. Use Skale to create an ETH sidechain3. Use Avax to send a chain and connect to Avax's P chain4. Send an EVM chain using Polygon Supernet 5. Using BAS to create a side chain based on BNB Chain6. Use OP Stack to create a Rollup Appchain7. Use Caldera to make a Rollup Appchain (essentially an Op Stack) 8. Use Zk-Sync to make an L3 (estimated to be visible this year) 9. At the time of writing, Arbitrum also released their Orbit, which is also an Opstack-like L3 infrastructure 10. There are also projects under development such as Opside, Stackr, and Sovereign SDK. If your chain is non-EVM, then you can 1. Use the Cosmos SDK as a chain, or be independent and self-reliant, or share ATOM security (ICS proposal just passed) 2. Use Substrate to send a chain, or bid for a Boca slot, or connect to the Octopus Network, or don't connect anyone, and stay quiet for a long time 3. Use Celestia's Rollkit to make a Rollup Appchain, DA uses Celestia, and payment options 4. Make a Rollup Appchain with Dymension5. Use Saga to make a Rollup Appchain6. Make an L3 based on Starkware (estimated to be visible this year) 7. There must be many more under development, so I didn't know or missed something I hadn't thought of — “The options are just a little bit too rich?!” Second, which apps are suitable for making Appchain? Remember the online article a few months ago about whether Uniswap should become an Appchain and various debates? Generally speaking, if Uniswap becomes an Appchain, it will have advantages such as transaction fees, token value capture, resistance to Mev, and resource monopoly, but at the same time, it has gained a Debuff with reduced user experience, security, and composability. In short, we haven't seen Uniswap's plans or “ambitions” Compound originally wanted to take Substrate was used as a chain, then abandoned. Now Compound and Aave V3 are multi-chain deployments. Currently, the possibility of becoming an Appchain doesn't seem very likely to be Curve. They should have never had this plan; dYdX, which has no dependency on composability, chose to run to Cosmos as an application chain. We should see this year in Q3. Perhaps after Luna, this is the most anticipated Appchain. At that time, dYdX will also show a path for many builders — if you take the lead on a racetrack and don't have high requirements for composability, then independent to make a “sovereign and high-performance” Appchain is Mak...

1229d agody zhang#Appchain #Arbitrum #BNB Chain #Curve #DeFi #Uniswap #ZK #gaming #hongkong

Long article in ten thousand characters - In my eyes, blockchain technology trends are evolving on various racetracks

Author of the original article: Lao Bai Original source: Mirror says that the current trends seen in various sectors recently and that may occur in the next 6-12 months will be divided into public chains, DeFi (Dex, lending, stablecoins, synthetic assets, etc.), NFT, GamFi, and storage sectors 1. First, let's talk about the new public chain trend mainly in the three categories of ETH+L2, Move+Diem, and Modularization 1. ETH+L2ETH, the order of the next few months to the next few years is this Merge (expected to be September-December) EIP4488 (expected to be the end of the year or the beginning of next year) - Reduce the cost of CallData by several times EIP4844 - Proto-Danksharding (next year) - Introduce a new transaction type Blob, replace CallData, and further reduce Rollup's fees Danksharding - (after 2-4 years) - Equivalent to Proto-DankSharding+ sharding ETH is fine, but everything is slow. 2. L2OP Arb - I know the pain point of L2 right now, that is, it's still not cheap enough, not fast enough (compared to Alt L1 or Polygon's sidechains like Solana), so while preparing Nitro's main network (using WASM architecture to greatly improve performance and reduce costs), it also has Nova (a sidechain based on AnyTrust, but there is Rollup as a backup, so it's much safer than a normal sidechain) Op - coin issuance is the biggest highlight, and the recently released Bedrock architecture. To make future multi-clients possible - not only clients that support OP, but also zk Yo Fuel - I used a new Sway language to get Swayswap. In the testnet demo, it was incredibly fast, and I don't know if this speed can be maintained after the main network is online. Fuel is worth mentioning that he uses the UTXO model. As we all know, UTXO has a performance advantage over the account model, but since there is no shared state, it is not suitable for complex contract designs. Fuel uses UTXO ID + Strict Access List (Strict Access List) technology to solve this problem ZK series zk-sync - announced the news that it will be on the main network within 100 days. It is worth noting that many dapps on zk-sync are native, not current Some Dapps have migrated to Polygon in the past -- which announced zk-EVMStarkware -- and also announced expectations for issuing coins. What Starkware is most anticipated is recursive rollups, which can be categorized into L3 or even L4 to achieve higher TPS and lower costs. Scroll - just announced Pre-Alpha, which claims to be the most EVM compatible zk-RollupAztec - got an Aztec Connect (privacy as a service), and finally got the privacy I wanted to see, that is, existing ETH L1 dApps can connect to Aztec through a bridge contract and SDK to get privacy protection. There is no fragmentation of liquidity, and no redeployment is required. Although the current implementation is very rudimentary, not all types of Dapps can be accessed, and it also depends on the team's ability to cooperate in PR, this is at least what I think privacy should look like. I never thought we needed any privacy coins, nor did we need a separate privacy chain; what we needed was a privacy plug-in! Finally, let's talk about ZK and concurrent processing. I just saw TWI sent by Mr. Min Dao. In the next few months, a large number of ZK departments will be launched, and then? It's definitely another Defi tri-axe, DEX, loan, stablecoin... the public chain's local currency+various DEX and loan projects are estimated to let everyone take over the market, and what is more important is the further fragmentation of liquidity. This fragmentation is still unresolved. DAMM, where LRC collaborates with Starkware, may be a solution, but first, this can only integrate the liquidity of ZK Rollup and ETH L1; third, the complexity of this plan is very high. Originally, MEV on a single chain was a problem. Now integrating multiple chains, adding asynchronicity, MEV, and impermanent loss will become more complicated. Anyway, I'm very serious... Close...

1468d agody zhang#AMPL #Atom #Bitcoin SV #BNT #COMP #CRV #DAI #DeFi #DOT #Fun #Gamefi #LEND #LRC #NFTs #SOL #Uniswap #USDC #USDT #Ethereum #Platform coins #Bitcoin

In-depth roundtable discussion: Layer 2 and the new public chain, who is responsible for ups and downs?

Source: What new breakthroughs will A&T CapitalL2 and the new public chain have in the future? What are the hidden wealth codes? Comparing the two rollup plans, one is OP and one is ZK. What would their position and pattern be in the market? Is the Orbitor bridge more secure than other bridges? Why aren't the Optimism and Arbitrum ecosystems so popular? What are future opportunities to detonate these two ecosystems? How should I design a layer 2 tokenomics? As an ordinary user or investor, how can I participate in Layer 2 to have better returns? Text #Q1 compares the two rollup plans. One is OP and one is ZK. What will their position and pattern in the market be? Tonight we invite a representative of the OP, a representative of ZK, and two speakers. Nina: Actually, some time ago, our company published a special article explaining why we chose OP technology to expand this track. From our company's point of view, it's OP. We think it will be cheaper in the long run, because in fact, if you want to do an addition today, 1 plus 1 equals 2, then OP actually only needs to do 1 plus 1 equals 2, but if it's ZK, you have to use complex cryptographic algorithms to prove that it is indeed 1 plus 1, which is actually equal to 2. So it actually consumes more unnecessary resources. So we chose OP in terms of efficiency. Then another thing is, we think ZK does have a certain advantage in core data now, and we acknowledge it, but we also know that in the future, Ethereum 2.0, EIP-8848, and EIP-4488 will solve this problem and narrow the gap between OP and ZK in terms of core data. So we don't think this will be an advantage that will last long. Looking further, I think ZK and OP will coexist for a while. They may even coexist forever, but I have no way of providing a very suitable scene right now. Right now, ZK really doesn't have the technical maturity to face the market, and smart contracts aren't working well on it. Now you can see a relatively successful case, which is actually a single scenario like Starkware, DYDX, or Immutable X. I believe ZK will definitely have a more suitable scenario when the technology is mature enough. If ZK matures to the point where the market is satisfied, we at Arbitrum may also switch to this technology. We don't think this is a competition between technology and technology; it's all about expansion. Whichever works better, the project will support its future. Luozhu: Whether it's ZK or OP, the main goal is to build a decentralized digital world. As for OP and ZK, we now use a different method of proof, but I don't think it's that limited. OP and ZK will probably coexist for a short to long period of time, including some projects that use zk proof technology on OP, such as PS1, which uses ZK internally to create an optimistic privacy chain rollup of the UTXO model. Vitalik himself said that OP may be superior in the short term, but in the long run it may be a state of ZK, but it will be a state of coexistence for a long time. Nina just mentioned core data, ZK performance issues, and EVM compatibility issues. You can experience zkSync's v2 test network, including contract deployment. In fact, we already have a very friendly development experience. For Scroll, we will launch the demo version of ZK EVM in July of this year. You can directly use the previous tools to do some testing and deployment. Compatibility with EVM is a significant improvement. The second performance issue with ZK includes generation time and verification time, which is expected to be solved in the long run. But OP, I believe that if ZK technology is mature enough, they will also learn from related technology. Let me put it simply, whether it's OP or zk rollup, they are doing Layer 2 expansion. Simply put, we just put some...

1529d agody zhang#Arbitrum #dYdX #layer 2 #optimism #Rollup #Tokenomics #Uniswap #WEB3 #Ethereum #New public chain #Bitcoin

No more