It seems that this time it is my turn to enter ChatGPT's grayscale test 😆 Plus accounts can experience GPT's official AI Agent. To reiterate my point, I also mentioned it in my live broadcast with Amber last night: 1. In the future, large model LLMs will become AI agents themselves, and then occupy the vast majority of the market with their own entrance advantages. 2. Third-party AI agents are a rare business with a marginal effect that does not decrease. If you keep using other large models as an AI agent, the more users you have, the higher your cost. 3. The cost of large models is a fraction of that of third-party agents. Such a competition is too difficult. 4. Third-party agents can only rely on exclusive data to obtain some moats, although burning money to buy data is still what these large-model Internet giants are best at.

Crypto ETFs are on the rise For example, Bitwise promotes an index consisting of a basket of 10 cryptocurrencies. (Although various CEXs have been pushed several times in the past N years without success) The current progress is that the SEC first approved it, and then instructed to postpone it, so it has not been listed. In addition, from this 10 index, you can see many clues. First of all, America first So you can see XRP, SOL, SUI, LINK, AVAX, and even DOT PS: And compared to SUI, Aptos, the head of the three best of Facebook, looks a bit behind. This is followed by the exclusion of exchange tokens. Unfortunately, BNB is actually an important part of the cryptocurrency market Finally, among PoW tokens, LTC is the only one that loves LTC but not BCH and Doge, which shows that PoW does not gain additional advantages. Grayscale's GDLC index is more conservative, with only 5 targets: But it still follows this rule: (1) America first (2) Platform currency is not accepted (3) PoW does not have any special advantages Of course, the situation is similar with Grayscale GDLC, which was first endorsed by the SEC and then suspended. According to the analysis of relevant people familiar with the matter, it seems that a member of the SEC opposed the proposal, which led to the suspension, and the reason for the opposition was that there needs to be a standard before these index ETFs can be launched.

Let's digresse with Jason's post. How is the Entry/Exit Queue Time Calculated for Ethereum Staking? In the past, Ethereum set this [Churn] metric to prevent the number of nodes participating in the consensus from changing too quickly, leading to unnecessary fluctuations (everything is mainly 😂 stable). This churn rate slowly increases with the increase in the overall staking scale, for example, the current maximum is 8 nodes per epoch, which is 256/epoch in the chart if calculated based on 32 $ETH per node. An epoch on Ethereum consists of 32 blocks, or 6.4 minutes, with a maximum of 8 nodes entering/exiting every 6.4 minutes. This is the technical reason for limiting the queue for Ethereum Staking in and out. What about the follow-up? So Ethereum pushed the Prague upgrade, allowing a single node to deposit ETH from 32 to 2048 $ETH →. It is to deal with this kind of "big hit" mortgage or unpledge. Imagine if we assume that 600,000 Ethereum belongs to several large users and each person is full from 32 to 2,048 nodes, then their withdrawal will be shortened from 9 days of queuing to about 3 hours. In other words, before the information ferments, the tokens have already been recharged to Binance OKX 😂 Of course, after the Prague upgrade, the adoption of large nodes is currently slowly increasing, and after looking at beaconcha, roughly speaking, about four or five hundred people have upgraded to large node mode. There will definitely be more and more large nodes on Ethereum in the future, and we will also see less and less queue time.

Today, I found a tutorial to keep it so that if someone studies EOS/Vaulta/$A in the future, it may be used. The EOS chain is quite complex, and the data documents are relatively scattered. First of all, EOS is now called Vaulta, and the token is called $A, which everyone knows. Amazingly, Vaulta now has two chains side by side: One is the traditional virtual machine of EOS, which is the familiar free transfer model with RAM mechanism. One is an EVM-compatible chain, using the Ethereum gas mechanism. The two chains are crossed through an official gateway. The cross-chain method is very distinctive, which is to transfer money directly to the gateway (eosio.evmin), and then note the memo address (0x1234...... 5678）。 What's even more amazing is that the token $EOS still exists today and can be swapped 1:1 with $A. The role of the old $EOS is to be used cross-chain. Presumably, since Vaulta had not changed its name when it launched EVM before, this cross-chain gateway can only cross $EOS. So it created the most magical situation: The token on Vaulta is A The token on the Vaulta EVM is EOS Although the two are 1:1 interchangeable 😂 In addition, a few details to note are: First of all, it's not about transferring A to the gateway, it's about transferring EOS to the gateway. How do I change my $EOS for $A? It is also a transfer, which is achieved directly by transferring money to core.vaulta In addition, the cross-chain fee is 0.1 EOS So below 0.1 EOS will fail across chains, and it takes several attempts to succeed.

Due to $SPK pull the plate Resulting in the interest rate of stablecoins in Sparkfi The cost performance goes up directly

Trump wants to replace the corn syrup (glucose fructose syrup) in Coca-Cola with real cane sugar, he really understands it too well 😂. One of the two double-blind experiments I would most like to do: 1. Can sucrose cola and glucose fructose syrup cola be drunk? 2. Pepsi and Coca-Cola, can you drink it? The level of unsolved mysteries in life.

What makes me happiest is that Ethereum has finally begun to [substantially] accelerate mainnet L1. Currently, the average Ethereum gas limit has climbed to 3700W and will soon increase to 4500W across the board, which means that the Ethereum mainnet is about 50% faster than before (3000W). After all, increasing the gas limit is the most immediate solution to improve Ethereum's performance. In fact, it is very easy to understand, you have to pay gas for every transaction you use; The maximum amount of gas that an Ethereum block can handle is the gas limit. For example, if I transfer $ETH to someone, the gas consumption is 21,000, and the gas limit of an Ethereum block is 3,700w, which means that within 12 seconds (one block interval), Ethereum can process up to 3,700w÷2.1w=1,761 transactions, and the theoretical peak TPS is equivalent to 146. Of course, because transferring ETH is the least gas-consuming method, complex DeFi transactions actually take up more gas. Therefore, the actual TPS of ETH is a bit lower. Many people ask, why doesn't Ethereum directly increase the gas limit higher? For example, the upper limit of 400 million mentioned earlier, wouldn't that be faster? You need to know that Ethereum as a blockchain – remember the nature of blockchain? That is, the ledger cannot be tampered with. The premise that the ledger cannot be tampered with is that there are enough copies of the ledger on the whole network. As we all know, there is no subsidy for maintaining a full node account. If the gas limit is blindly raised, the blockchain ledger will become larger, so that the burden on nodes will become heavier and heavier, and then the number of ledgers on the entire network will decrease. The previous archive node has reached 20 TB, to be honest, who always has a 20 TB large hard drive for charity, so I really can't increase the gas limit in the past. Because once the number of ledgers is reduced, or household devices cannot participate in consensus, this is "shaking the chain". So why did Ethereum easily increase to 3700w this time? The reason is that Geth recently released the v1.16 client, which finally caused Ethereum's archive nodes to plummet from 20 TB to 1.9 TB, a significant reduction of 90%, so that the foundation and nodes have the courage to continue to raise the gas limit. So why a 10x drop? This is because archive nodes used hash-based storage mode in the past, but now they have changed to path-based storage mode (PBSS). In the past hash mode, the main focus of the archive node was to "save as much as possible", storing the complete history tree version. No automatic trimming; All data is stacked directly in the primary database, resulting in a high degree of redundancy in historical data. After changing to a path-based storage mode, nodes are indexed in the tree through paths. Instead of mindlessly storing all values, only the difference values, the volume can be very compact and the duplicate content is greatly reduced. In addition, only one full state snapshot is persisted in the database, corresponding to the full state of the last 128 blocks. And pruning data from 90,000 blocks (12.5 days) ago, which avoids storing a full snapshot of all historical blocks, greatly reducing redundancy. Therefore, it is precisely because the size of the archive node has been reduced, and even an HDD can run, that the Ethereum mainnet dares to directly increase the gas limit. According to the current method, it is not a problem to mention 4500w in an orderly and slow manner, and then mention the upper limit of 100 million. At that time, it will not be difficult for the TPS of Ethereum mainnet L1 to easily exceed 100, and as the "world stablecoin settlement layer" at that time, it is still quite sufficient. The most rebellious thing is that considering the current sky-defying gas fee of Tron (in non-energy mode), and even in the near future, we can see that Ethereum's fees are lower 😂 than TRON

Seeing the news that @yalaorg Yala's TGE will be selected next week, it is the time when BTC has just reached a new high, the second cake has climbed to the peak, liquidity is abundant, and the mood is FOMO. Looking back at Babylon's TGE time at that time, it was accidentally chosen in mid-April, which was the only time BTC fell below 80K this year. It is really time and fate. Someone once asked what is the difference between the two being BTC Fi, and today I opened a post on the eve of TGE to talk about it. First, the way the protocol is implemented. Babylon will have its own independent L1 in the future, using Bitcoin collateral as additional security, somewhat similar to Bitcoin sidechains. Specifically, $BABY holders run PoS nodes and are responsible for producing blocks. The Bitcoin staker is responsible for verifying and making the final signature. The advantage of this is that independent L1 has faster performance and greater freedom. Yala, on the other hand, directly uses Ethereum as the settlement layer and does not issue its own chain, which is equivalent to a wrapped version of Bitcoin. In this way, chain security is guaranteed by Ethereum, and the bottlenecks of protocol security are mainly focused on cross-chain bridges and multi-signature. The benefit is that adopting Ethereum will result in better compatibility and, obviously, faster product development. The product goals of the other two are also different. After Babylon has its own L1, it will focus on DeFi, RWA, and various applications like other public chains. The main thing is a big and complete. After Yala crossed BTC to Ethereum, it focused on the stablecoin track. It allows $YBTC to overcollateralize to generate stablecoin $YU (somewhat similar to ETH overcollateralization in MakerDAO to generate DAI/USDS), so $YALA specializes in stablecoins. Considering the recent smooth progress of the Genius Act, it is now timely to borrow the east wind. Now more and more people are realizing that the killer applications of blockchain are actually Bitcoin, Ethereum and stablecoins, and the applications built around them hope to be the icing on the cake.

The best time to plant a tree is ten years ago The second is now If you are interested in ETH micro strategies Welcome to contact 👇 us

AI is influencing Web3, will AgentFi be the core variable of the next wave? Tomorrow night, we will join hands with industry alpha to have an in-depth discussion on "AgentFi: New Opportunities in the Web3 Intelligent Economy under the AI Wave": Event details🔽 Date: July 18 (Friday) Time: 20:00 Language: Chinese Guest lineup: - Haotian @tmel0211 (Independent Researcher) - 0xTodd @0x_Todd (Partner at Nothing Research / Co-founder of Ebunker) - LaoBai @wuhuoqiu (Former Partner in Investment and Research at ABCDE) - Howe @0xcryptoHowe (Intern at Amber Group) - Anne @AnneXingxb (AI Researcher at Amber Group) Topics include entrepreneurial opportunities in Agent, structural variables, and the integration path and future outlook of AI + Crypto. Welcome to listen and foresee the next explosion point of the intelligent economy together👉 See you tomorrow night!