7*24

Hot topic

Market

Activities

On April 20, LayerZero Labs released a statement regarding an attack incident, stating that on April 18, KelpDAO was attacked, resulting in a loss of approximately $290 million. Preliminary assessments suggest that the attackers are linked to the North Korean group Lazarus Group, specifically the TraderTraitor faction. The attack was executed by poisoning the downstream RPC infrastructure relied upon by its decentralized validation network (DVN). The attackers controlled certain RPC nodes and conducted a DDoS attack to lure the system into switching to malicious nodes, thereby fabricating cross-chain transactions. All affected RPC nodes have been taken offline and replaced, and the DVN has resumed operations. LayerZero emphasized that this incident was limited to the rsETH application configuration of KelpDAO and did not impact other assets or applications. The reason for this is that KelpDAO was using a single DVN (1/1) architecture at the time and did not implement the multi-DVN redundancy mechanism that is recommended by the official guidelines, resulting in a lack of independent validation nodes to identify forged messages. LayerZero pointed out that there were no vulnerabilities in its protocol, and applications with multi-DVN configurations were not affected, indicating that there is no contagious risk within the system.

[LayerZero Reports KelpDAO Theft Incident: Confirms Only rsETH Configuration Affected]On April 20, LayerZero Labs released a statement regarding an attack incident, stating that on April 18, KelpDAO was attacked, resulting in a loss of approximately $290 million. Preliminary assessments suggest that the attackers are linked to the North Korean group Lazarus Group, specifically the TraderTraitor faction. The attack was executed by poisoning the downstream RPC infrastructure relied upon by its decentralized validation network (DVN). The attackers controlled certain RPC nodes and conducted a DDoS attack to lure the system into switching to malicious nodes, thereby fabricating cross-chain transactions. All affected RPC nodes have been taken offline and replaced, and the DVN has resumed operations. LayerZero emphasized that this incident was limited to the rsETH application configuration of KelpDAO and did not impact other assets or applications. The reason for this is that KelpDAO was using a single DVN (1/1) architecture at the time and did not implement the multi-DVN redundancy mechanism that is recommended by the official guidelines, resulting in a lack of independent validation nodes to identify forged messages. LayerZero pointed out that there were no vulnerabilities in its protocol, and applications with multi-DVN configurations were not affected, indicating that there is no contagious risk within the system.

LayerZero Reports KelpDAO Theft Incident: Confirms Only rsETH Configuration Affected

On April 20, Lu Liping, a professor and doctoral supervisor at the School of Finance and Economics of Renmin University of China, stated that the long-term hedging function of gold has not been lost. However, there are some short-term fluctuations. For instance, following the US-Iran conflict, oil prices surged, leading to an increase in inflation expectations. This rise in inflation expectations has resulted in the Federal Reserve not lowering interest rates, or potentially raising them, which in turn increases the holding costs of gold. Nevertheless, in the long term, the demand for gold as a hedge remains very strong. (China Business Journal)

[Lu Liping: The Long-term Hedging Function of Gold Remains Intact, Short-term Volatility Due to Various Factors]On April 20, Lu Liping, a professor and doctoral supervisor at the School of Finance and Economics of Renmin University of China, stated that the long-term hedging function of gold has not been lost. However, there are some short-term fluctuations. For instance, following the US-Iran conflict, oil prices surged, leading to an increase in inflation expectations. This rise in inflation expectations has resulted in the Federal Reserve not lowering interest rates, or potentially raising them, which in turn increases the holding costs of gold. Nevertheless, in the long term, the demand for gold as a hedge remains very strong. (China Business Journal)

Lu Liping: The Long-term Hedging Function of Gold Remains Intact, Short-term Volatility Due to Various Factors

On April 20, analysts stated that the shipping through the Strait of Hormuz has been intermittent, highlighting the uncertain status of this globally significant oil and gas chokepoint. However, one thing is clear: even if the fighting ceases, it may take months—possibly even years—for oil flow through this narrow waterway to return to pre-war levels. Analysts believe that the pace of recovery depends not only on diplomatic progress between Washington and Tehran but also on logistics conditions, the availability of tanker insurance, freight rates, and whether shipowners are willing to take the risk of passage. It is expected that the complete rebalancing of the global tanker fleet and the restoration of Gulf loading operations to pre-war rhythms will be uneven, potentially requiring at least eight to twelve weeks even under favorable conditions.

[Analysts: Reopening the Strait of Hormuz is Feasible, but Full Restoration of Oil Flow May Take Years]On April 20, analysts stated that the shipping through the Strait of Hormuz has been intermittent, highlighting the uncertain status of this globally significant oil and gas chokepoint. However, one thing is clear: even if the fighting ceases, it may take months—possibly even years—for oil flow through this narrow waterway to return to pre-war levels. Analysts believe that the pace of recovery depends not only on diplomatic progress between Washington and Tehran but also on logistics conditions, the availability of tanker insurance, freight rates, and whether shipowners are willing to take the risk of passage. It is expected that the complete rebalancing of the global tanker fleet and the restoration of Gulf loading operations to pre-war rhythms will be uneven, potentially requiring at least eight to twelve weeks even under favorable conditions.

Analysts: Reopening the Strait of Hormuz is Feasible, but Full Restoration of Oil Flow May Take Years

The Cashmere team, behind the LayerZero-based anti-MEV DEX aggregator, announced via Twitter that they have open-sourced the Cashmere Agent SDK. This SDK is designed as an execution layer specifically for AI Agents, allowing them to perform cross-chain execution operations directly. Core features include simulating cross-chain routing, estimating gas, slippage, and execution time, selecting the optimal execution path, and executing transactions directly on-chain. The current version supports mainstream frameworks such as LangChain, OpenAI tool calling, and crewAI, and includes features like the simulate() engine, cross-chain USDC transfers, and Aptos 64-byte key support.

[Cashmere Labs Open Sources Agent SDK for Cross-Chain Execution of AI Agents]The Cashmere team, behind the LayerZero-based anti-MEV DEX aggregator, announced via Twitter that they have open-sourced the Cashmere Agent SDK. This SDK is designed as an execution layer specifically for AI Agents, allowing them to perform cross-chain execution operations directly. Core features include simulating cross-chain routing, estimating gas, slippage, and execution time, selecting the optimal execution path, and executing transactions directly on-chain. The current version supports mainstream frameworks such as LangChain, OpenAI tool calling, and crewAI, and includes features like the simulate() engine, cross-chain USDC transfers, and Aptos 64-byte key support.

Cashmere Labs Open Sources Agent SDK for Cross-Chain Execution of AI Agents

According to monitoring by Dongcha Beating, xAI and Tesla CEO Elon Musk responded on X when asked which version of Grok would achieve AGI, stating only "Grok 5." He directly locked the AGI milestone to this version. He subsequently shared the complete version roadmap: 4.4=1T, 4.5=1.5T, 4.6=3T, 4.7=6T, 4.8=10T, 4.9=???, 5.0=AGI, 6.0=ASI, 7.0=ASI2. Currently, the version available for testing is 4.3 beta (0.5T). According to this roadmap, the parameters for versions 4.4 to 4.8 scale up to 10T, corresponding to his previous statement about releasing an improved base model every two weeks; 4.9 is left blank, while 5.0 and beyond are directly labeled AGI and ASI (Artificial Superintelligence, a hypothetical stage after AGI where intelligence levels exceed human capabilities), and ASI2. At the end of this complete roadmap, Musk included a shrug and a laughing-crying emoji, suggesting that the labels after 5.0 are somewhat joking and not commitments to parameters or capabilities; however, the earlier statement "Grok 5 = AGI" was made without emojis, indicating his public stance on the timeline for AGI.

[Musk: Grok 5 Will Achieve AGI, Reveals Complete Roadmap from 4.4 to 7.0]According to monitoring by Dongcha Beating, xAI and Tesla CEO Elon Musk responded on X when asked which version of Grok would achieve AGI, stating only "Grok 5." He directly locked the AGI milestone to this version. He subsequently shared the complete version roadmap: 4.4=1T, 4.5=1.5T, 4.6=3T, 4.7=6T, 4.8=10T, 4.9=???, 5.0=AGI, 6.0=ASI, 7.0=ASI2. Currently, the version available for testing is 4.3 beta (0.5T). According to this roadmap, the parameters for versions 4.4 to 4.8 scale up to 10T, corresponding to his previous statement about releasing an improved base model every two weeks; 4.9 is left blank, while 5.0 and beyond are directly labeled AGI and ASI (Artificial Superintelligence, a hypothetical stage after AGI where intelligence levels exceed human capabilities), and ASI2. At the end of this complete roadmap, Musk included a shrug and a laughing-crying emoji, suggesting that the labels after 5.0 are somewhat joking and not commitments to parameters or capabilities; however, the earlier statement "Grok 5 = AGI" was made without emojis, indicating his public stance on the timeline for AGI.

Musk: Grok 5 Will Achieve AGI, Reveals Complete Roadmap from 4.4 to 7.0

On April 20, according to official news from MEET48, the organization will co-host the Hong Kong Web3 Carnival Side Event "RWA Night in Entertainment" tonight from 19:00 to 23:00 with Hash Global and Nine Lives Society, marking a highlight moment for RWA in Hong Kong alongside projects like POCHAIN, OLAXBT, TATTO, IPDEX, and Renaiss. Additionally, MEET48 COO Kai Xu will deliver a keynote speech tomorrow morning from 10:45 to 11:00 at the main venue of the Hong Kong Web3 Carnival, sharing forward-looking strategies on the intersection of AI and Crypto in the entertainment industry. As the world's first idol fan economy ecosystem built on Web3, deeply integrating AI and UGC, MEET48 is developing a platform focused on virtual idols and AI-Web5 entertainment. Previously, the Hong Kong Web3 Carnival announced MEET48 as a diamond sponsor, with their booth located at HKCEC, 5BCDE, B03, from April 20 to April 23. Don't miss it!

[MEET48 Hosts Hong Kong Web3 Carnival Side Event Tonight, Igniting RWA Night in Entertainment; COO to Speak Tomorrow Morning]On April 20, according to official news from MEET48, the organization will co-host the Hong Kong Web3 Carnival Side Event "RWA Night in Entertainment" tonight from 19:00 to 23:00 with Hash Global and Nine Lives Society, marking a highlight moment for RWA in Hong Kong alongside projects like POCHAIN, OLAXBT, TATTO, IPDEX, and Renaiss. Additionally, MEET48 COO Kai Xu will deliver a keynote speech tomorrow morning from 10:45 to 11:00 at the main venue of the Hong Kong Web3 Carnival, sharing forward-looking strategies on the intersection of AI and Crypto in the entertainment industry. As the world's first idol fan economy ecosystem built on Web3, deeply integrating AI and UGC, MEET48 is developing a platform focused on virtual idols and AI-Web5 entertainment. Previously, the Hong Kong Web3 Carnival announced MEET48 as a diamond sponsor, with their booth located at HKCEC, 5BCDE, B03, from April 20 to April 23. Don't miss it!

MEET48 Hosts Hong Kong Web3 Carnival Side Event Tonight, Igniting RWA Night in Entertainment; COO to Speak Tomorrow Morning

According to Beating monitoring, Google is in talks with Marvell Technology to collaborate on the development of two custom chips designed to enhance the efficiency of AI model operations. This move signifies Google's acceleration in diversifying its AI hardware suppliers to reduce reliance on its long-time partner Broadcom and cut costs. The two chips include a new type of Tensor Processing Unit (TPU) specifically for AI inference, and a Memory Processing Unit (MPU) that will work alongside existing TPUs. Google plans to produce approximately 2 million MPUs, with design work expected to be completed as early as next year. Inference chips are primarily responsible for running trained models and are central to driving commercial products like automated agents. Currently, Broadcom is Google's sole long-term design partner for TPUs and charges high fees for each chip produced. With the surge in TPU demand, Google is mitigating risks by bringing in Marvell and MediaTek.

[Google Partners with Marvell to Develop Two AI Inference Chips, Aiming to Reduce Dependence on Broadcom]According to Beating monitoring, Google is in talks with Marvell Technology to collaborate on the development of two custom chips designed to enhance the efficiency of AI model operations. This move signifies Google's acceleration in diversifying its AI hardware suppliers to reduce reliance on its long-time partner Broadcom and cut costs. The two chips include a new type of Tensor Processing Unit (TPU) specifically for AI inference, and a Memory Processing Unit (MPU) that will work alongside existing TPUs. Google plans to produce approximately 2 million MPUs, with design work expected to be completed as early as next year. Inference chips are primarily responsible for running trained models and are central to driving commercial products like automated agents. Currently, Broadcom is Google's sole long-term design partner for TPUs and charges high fees for each chip produced. With the surge in TPU demand, Google is mitigating risks by bringing in Marvell and MediaTek.

Google Partners with Marvell to Develop Two AI Inference Chips, Aiming to Reduce Dependence on Broadcom

On April 20, the Monetary Authority of Singapore (MAS) urged banks to enhance their cybersecurity defenses due to security concerns arising from the spread of the Mythos AI model by Anthropic in Asia.

[MAS Urges Banks to Strengthen Cybersecurity Defenses]On April 20, the Monetary Authority of Singapore (MAS) urged banks to enhance their cybersecurity defenses due to security concerns arising from the spread of the Mythos AI model by Anthropic in Asia.

MAS Urges Banks to Strengthen Cybersecurity Defenses

On April 20, according to SoSoValue data, the US spot Bitcoin ETF recorded a net inflow of $996 million last week.

[US Spot Bitcoin ETF Sees $996 Million Net Inflow Last Week]On April 20, according to SoSoValue data, the US spot Bitcoin ETF recorded a net inflow of $996 million last week.

US Spot Bitcoin ETF Sees $996 Million Net Inflow Last Week

according to the on-chain analyst Yu Jin's monitoring, the ancient whale who has held 80,000 bitcoins for 14 years has once again transferred 10,000 BTC to the same address.

[The ancient whale who held 80,000 Bitcoins for 14 years transferred another 10,000 BTC to the same address]according to the on-chain analyst Yu Jin's monitoring, the ancient whale who has held 80,000 bitcoins for 14 years has once again transferred 10,000 BTC to the same address.

The ancient whale who held 80,000 Bitcoins for 14 years transferred another 10,000 BTC to the same address

Bitcoin Created By The CIA? Chinese Professor Jiang Xueqin Makes Bold Claim

BIT Research: Five Signals Flash Simultaneously, Has the Bitcoin Bear Market Ended?

Bitcoin Recovery Fails To Lift Market Sentiment From Extreme Fear

Crypto And Financial Industry Giants Reveal What X Money Launch Means

Analyst Exposes Bitcoin Market Maker Buy Strategy, Shows What Happens When Accumulation Ends

Bitcoin Mining Network Collapsing Into AI At Record Pace, Analyst Warns

Conversation with Bitwise Advisor: From K-Shaped Economy to AI Taking Jobs, How Can Bitcoin Save the Younger Generation?

Analyst Reveals The Chances Of Bitcoin Price Crashing Again

$20M Crypto Scam Lands Texas Man 23-Year Prison Sentence

Bitcoin Miners Selling Nears Exhaustion – What Comes Next

Article

Analysis: Divergence Between Bitcoin's Price Action and Realized Profit Indicates Possible Peak in Rebound

The ancient whale who held 80,000 Bitcoins for 14 years transferred another 10,000 BTC to the same address

Descargo de responsabilidad：El contenido anterior no representa las posiciones de HTX.，HTX no ofrece ninguna recomendación de trading.。

Artículos relacionados

Todos los comentarios0Lo más recientePopular

Temas populares

Artículos relacionados