7*24

Hot topic

Market

Activities

On July 6, prediction platform Polymarket reported that the France team holds the highest implied championship probability at 35.4%. French player Kylian Mbappé leads the World Cup Golden Boot market with odds of 51%.

[Polymarket: France Team Tops Championship Probability]On July 6, prediction platform Polymarket reported that the France team holds the highest implied championship probability at 35.4%. French player Kylian Mbappé leads the World Cup Golden Boot market with odds of 51%.

Polymarket: France Team Tops Championship Probability

On July 6, August Hyldgaard of the Danish Bank stated in a report that after taking its first action in over three years in June, the Danish central bank may intervene further to support the Danish krone. Data released on Thursday showed that the central bank net purchased 700 million kroner in the market to defend its peg with the euro. Hyldgaard noted that history indicates the central bank will respond to the need for expanded intervention, with the scale of intervention potentially around 50 billion kroner. The Danish Bank expects the central bank to set a short-term euro cap at 7.4758 kroner. Data from the London Stock Exchange Group showed the euro was flat against the Danish krone at 7.4746 kroner, after previously reaching an 11-year high of 7.4759 kroner on June 25.

[Danish Bank: Central Bank May Take Further Action to Support Danish Krone]On July 6, August Hyldgaard of the Danish Bank stated in a report that after taking its first action in over three years in June, the Danish central bank may intervene further to support the Danish krone. Data released on Thursday showed that the central bank net purchased 700 million kroner in the market to defend its peg with the euro. Hyldgaard noted that history indicates the central bank will respond to the need for expanded intervention, with the scale of intervention potentially around 50 billion kroner. The Danish Bank expects the central bank to set a short-term euro cap at 7.4758 kroner. Data from the London Stock Exchange Group showed the euro was flat against the Danish krone at 7.4746 kroner, after previously reaching an 11-year high of 7.4759 kroner on June 25.

Danish Bank: Central Bank May Take Further Action to Support Danish Krone

On July 6, according to the Cyberspace Administration, the 'Clear and Bright: Rectifying AI Application Chaos' initiative has been underway since April 2026. The Central Cyberspace Administration has focused on issues such as failure to comply with large model registration obligations, insufficient security and review capabilities of AI platforms, AI data poisoning, and inadequate implementation of synthetic content identification. Local cyberspace departments have been tasked with advancing the key rectification efforts of the first phase and have established a 'Reporting Zone for AI Application Chaos' to specifically handle public complaints. Through collaborative efforts, over 14,000 non-compliant websites, applications, and AI products have been addressed in the first phase, over 6 million pieces of illegal information have been cleared, more than 26,000 accounts have been dealt with, and over 1,300 non-compliant AI products and 9 non-compliant open-source datasets have been removed, achieving positive progress in all areas.

[Central Cyberspace Administration Launches First Phase of 'Clear and Bright: Rectifying AI Application Chaos' Initiative]On July 6, according to the Cyberspace Administration, the 'Clear and Bright: Rectifying AI Application Chaos' initiative has been underway since April 2026. The Central Cyberspace Administration has focused on issues such as failure to comply with large model registration obligations, insufficient security and review capabilities of AI platforms, AI data poisoning, and inadequate implementation of synthetic content identification. Local cyberspace departments have been tasked with advancing the key rectification efforts of the first phase and have established a 'Reporting Zone for AI Application Chaos' to specifically handle public complaints. Through collaborative efforts, over 14,000 non-compliant websites, applications, and AI products have been addressed in the first phase, over 6 million pieces of illegal information have been cleared, more than 26,000 accounts have been dealt with, and over 1,300 non-compliant AI products and 9 non-compliant open-source datasets have been removed, achieving positive progress in all areas.

Central Cyberspace Administration Launches First Phase of 'Clear and Bright: Rectifying AI Application Chaos' Initiative

According to DeFiLlama data, as of July 6, Huobi HTX has achieved a net capital inflow of over $11 million in the past 24 hours, ranking second among mainstream centralized exchanges (CEX) globally. It is reported that Huobi HTX has been continuously promoting product innovation and brand operations, recently launching a TradFi U.S. stock section and rolling out a series of business activities around the World Cup, further boosting user activity and capital inflow on the platform.

[DeFiLlama: Huobi HTX Sees Over $11 Million Net Capital Inflow in 24 Hours, Ranking Second Among Global CEXs]According to DeFiLlama data, as of July 6, Huobi HTX has achieved a net capital inflow of over $11 million in the past 24 hours, ranking second among mainstream centralized exchanges (CEX) globally. It is reported that Huobi HTX has been continuously promoting product innovation and brand operations, recently launching a TradFi U.S. stock section and rolling out a series of business activities around the World Cup, further boosting user activity and capital inflow on the platform.

DeFiLlama: Huobi HTX Sees Over $11 Million Net Capital Inflow in 24 Hours, Ranking Second Among Global CEXs

On July 6, it was announced that the International Conference on Machine Learning (ICML) 2026 was held in Seoul, South Korea, where the annual award-winning papers were revealed. The paper by Tsinghua University's Huang Gao team in collaboration with Alibaba, titled 'The Flexibility Trap: Rethinking the Value of Arbitrary Order in Diffusion Language Models,' won the Outstanding Paper Award. The research reveals that the flexibility of arbitrary generation order in diffusion language models limits the model's potential in general reasoning tasks such as mathematics and programming. By abandoning arbitrary order and adopting the traditional left-to-right generation, the method not only becomes simpler but also significantly improves reasoning accuracy. Another Outstanding Paper Award was given to researchers from MIT and Yale University for their study proposing a high-accuracy sampling algorithm for diffusion models (High-Accuracy Sampling for Diffusion Models and Log-Concave Distributions), achieving exponential optimization in the number of steps (or sampling complexity) required to reach the target sampling accuracy. Additionally, a position paper titled 'Position: The Alignment Community is Unintentionally Building a Censor’s Toolkit,' co-authored by researchers from Ludwig Maximilian University of Munich and independent researchers, highlighted the dual-use risks of current AI alignment technologies, which can easily be manipulated into tools for information censorship. The Time-Tested Award was presented to the Google DeepMind team for their classic reinforcement learning algorithm 'Asynchronous Methods for Deep Reinforcement Learning,' published in 2016. This research introduced the asynchronous advantage actor-critic (A3C) architecture, significantly enhancing the training efficiency of deep reinforcement learning and marking the beginning of an era where intelligent agents can be efficiently trained using standard multi-core CPUs.

[Tsinghua Huang Gao Team Wins Outstanding Paper Award at ICML 2026; Time-Tested Award Goes to Classic Algorithm A3C]On July 6, it was announced that the International Conference on Machine Learning (ICML) 2026 was held in Seoul, South Korea, where the annual award-winning papers were revealed. The paper by Tsinghua University's Huang Gao team in collaboration with Alibaba, titled 'The Flexibility Trap: Rethinking the Value of Arbitrary Order in Diffusion Language Models,' won the Outstanding Paper Award. The research reveals that the flexibility of arbitrary generation order in diffusion language models limits the model's potential in general reasoning tasks such as mathematics and programming. By abandoning arbitrary order and adopting the traditional left-to-right generation, the method not only becomes simpler but also significantly improves reasoning accuracy. Another Outstanding Paper Award was given to researchers from MIT and Yale University for their study proposing a high-accuracy sampling algorithm for diffusion models (High-Accuracy Sampling for Diffusion Models and Log-Concave Distributions), achieving exponential optimization in the number of steps (or sampling complexity) required to reach the target sampling accuracy. Additionally, a position paper titled 'Position: The Alignment Community is Unintentionally Building a Censor’s Toolkit,' co-authored by researchers from Ludwig Maximilian University of Munich and independent researchers, highlighted the dual-use risks of current AI alignment technologies, which can easily be manipulated into tools for information censorship. The Time-Tested Award was presented to the Google DeepMind team for their classic reinforcement learning algorithm 'Asynchronous Methods for Deep Reinforcement Learning,' published in 2016. This research introduced the asynchronous advantage actor-critic (A3C) architecture, significantly enhancing the training efficiency of deep reinforcement learning and marking the beginning of an era where intelligent agents can be efficiently trained using standard multi-core CPUs.

Tsinghua Huang Gao Team Wins Outstanding Paper Award at ICML 2026; Time-Tested Award Goes to Classic Algorithm A3C

On July 6, a comparison between U.S. stocks and South Korean tech stocks reveals a distinct 'leverage ETF dominance' structure among leading South Korean semiconductor firms such as SK Hynix and Samsung Electronics. The total asset scale of related single-stock ETFs and associated products has multiplied several times compared to the average daily trading volume of the underlying stocks. In contrast, U.S. tech stocks like Micron Technology, Tesla, and Nvidia primarily engage in spot trading, with ETF sizes significantly lower than their average daily trading volumes. Data shows (as of June 29, 2026): the total assets of SK Hynix's single-stock ETFs and related products reached $19.04 billion, while its average daily trading volume was only $4.47 billion; for Samsung Electronics, the corresponding figures were $12.43 billion and $4.49 billion. In comparison, Micron Technology's ETF size was $9.88 billion against an average daily trading volume of $27.47 billion; Tesla's was $5.95 billion to $23.56 billion; and Nvidia's was $5.57 billion to $28.75 billion—demonstrating a structure opposite to that of South Korean stocks. Due to the lack of individual stock options in the Korean market and a high proportion of retail investors, leveraged funds are highly concentrated in ETF products, causing stock prices to be more driven by passive rebalancing actions. When the market declines, ETFs are forced to sell, but with insufficient market absorption capacity, this easily amplifies price fluctuations, creating a unique 'amplifier mechanism' in the South Korean semiconductor market.

[Leverage Concentration and High Retail Investor Proportion: South Korea's Semiconductor Market 'Amplifier Mechanism' Draws Attention]On July 6, a comparison between U.S. stocks and South Korean tech stocks reveals a distinct 'leverage ETF dominance' structure among leading South Korean semiconductor firms such as SK Hynix and Samsung Electronics. The total asset scale of related single-stock ETFs and associated products has multiplied several times compared to the average daily trading volume of the underlying stocks. In contrast, U.S. tech stocks like Micron Technology, Tesla, and Nvidia primarily engage in spot trading, with ETF sizes significantly lower than their average daily trading volumes. Data shows (as of June 29, 2026): the total assets of SK Hynix's single-stock ETFs and related products reached $19.04 billion, while its average daily trading volume was only $4.47 billion; for Samsung Electronics, the corresponding figures were $12.43 billion and $4.49 billion. In comparison, Micron Technology's ETF size was $9.88 billion against an average daily trading volume of $27.47 billion; Tesla's was $5.95 billion to $23.56 billion; and Nvidia's was $5.57 billion to $28.75 billion—demonstrating a structure opposite to that of South Korean stocks. Due to the lack of individual stock options in the Korean market and a high proportion of retail investors, leveraged funds are highly concentrated in ETF products, causing stock prices to be more driven by passive rebalancing actions. When the market declines, ETFs are forced to sell, but with insufficient market absorption capacity, this easily amplifies price fluctuations, creating a unique 'amplifier mechanism' in the South Korean semiconductor market.

Leverage Concentration and High Retail Investor Proportion: South Korea's Semiconductor Market 'Amplifier Mechanism' Draws Attention

On July 6, Oracle (ORCL.US) saw its shares rise nearly 4% in pre-market trading, reaching $145.81. Recently, Oracle announced the launch of four new Fusion intelligent applications designed to help organizations improve inventory visibility, reduce the impact of supplier and operational issues, and enhance manufacturing efficiency, thereby improving supply chain performance. These new intelligent applications are integrated within Oracle Fusion Cloud Supply Chain & Manufacturing (SCM) and are driven by a collaborative team of multiple specialized AI agents, featuring result-oriented, proactive response, and logical reasoning capabilities, specifically designed for enterprise-level applications. Additionally, to further enhance organizational supply chain resilience, Oracle has also introduced a new inventory optimization feature.

[Oracle Shares Rise Nearly 4% Pre-Market with Launch of Four New Fusion Intelligent Applications]On July 6, Oracle (ORCL.US) saw its shares rise nearly 4% in pre-market trading, reaching $145.81. Recently, Oracle announced the launch of four new Fusion intelligent applications designed to help organizations improve inventory visibility, reduce the impact of supplier and operational issues, and enhance manufacturing efficiency, thereby improving supply chain performance. These new intelligent applications are integrated within Oracle Fusion Cloud Supply Chain & Manufacturing (SCM) and are driven by a collaborative team of multiple specialized AI agents, featuring result-oriented, proactive response, and logical reasoning capabilities, specifically designed for enterprise-level applications. Additionally, to further enhance organizational supply chain resilience, Oracle has also introduced a new inventory optimization feature.

Oracle Shares Rise Nearly 4% Pre-Market with Launch of Four New Fusion Intelligent Applications

On July 6, executives from several tech companies recently adjusted their statements regarding the impact of AI on employment, shifting from previously emphasizing that "AI will massively replace jobs" to believing that AI is more about enhancing productivity and creating new positions. OpenAI CEO Sam Altman stated that the industry had previously underestimated the role of humans within AI systems; Anthropic CEO Dario Amodei noted that AI could lead to layoffs but could also help companies accomplish more business without increasing headcount, with outcomes depending on management decisions. A recent EY-Parthenon survey showed that the proportion of corporate executives believing that AI investment would lead to massive layoffs has decreased from 46% in January 2025 to 20% in May 2026. Joint research by Ramp and Revelio Labs indicated that companies with the highest AI investment levels experienced employee growth rates about 10% higher than similar companies without AI initiatives. Although Meta CEO Mark Zuckerberg and Amazon CEO Andy Jassy both stated that AI is expected to create more job opportunities, both companies have continued to streamline their organizations this year, raising market skepticism about their AI narratives. A survey by tech consulting firm Emergn revealed that about 20% of U.S. business managers reported that internal AI project reports tend to exaggerate effectiveness and downplay issues. Industry insiders believe that the commercialization of AI is progressing slower than market expectations, and the application effects across different industries and scenarios still show significant variation.

[AI Job Displacement Narrative Cools as Tech CEOs Shift to Optimism]On July 6, executives from several tech companies recently adjusted their statements regarding the impact of AI on employment, shifting from previously emphasizing that "AI will massively replace jobs" to believing that AI is more about enhancing productivity and creating new positions. OpenAI CEO Sam Altman stated that the industry had previously underestimated the role of humans within AI systems; Anthropic CEO Dario Amodei noted that AI could lead to layoffs but could also help companies accomplish more business without increasing headcount, with outcomes depending on management decisions. A recent EY-Parthenon survey showed that the proportion of corporate executives believing that AI investment would lead to massive layoffs has decreased from 46% in January 2025 to 20% in May 2026. Joint research by Ramp and Revelio Labs indicated that companies with the highest AI investment levels experienced employee growth rates about 10% higher than similar companies without AI initiatives. Although Meta CEO Mark Zuckerberg and Amazon CEO Andy Jassy both stated that AI is expected to create more job opportunities, both companies have continued to streamline their organizations this year, raising market skepticism about their AI narratives. A survey by tech consulting firm Emergn revealed that about 20% of U.S. business managers reported that internal AI project reports tend to exaggerate effectiveness and downplay issues. Industry insiders believe that the commercialization of AI is progressing slower than market expectations, and the application effects across different industries and scenarios still show significant variation.

AI Job Displacement Narrative Cools as Tech CEOs Shift to Optimism

On July 6, it was reported that major AI model applications such as Doubao under ByteDance, Tongyi Qianwen from Alibaba, and Yuanbao from Tencent are gradually suspending or adjusting their custom 'AI Companion' features. Doubao has notified users that it will discontinue the custom character feature on July 15 and will guide users in need to independent companion applications. This move aims to comply with new generative AI regulations in China that will take effect in mid-July. The regulations, promoted by the National Internet Information Office, focus on preventing the potential negative impacts of anthropomorphic AI services. They explicitly prohibit platforms from generating content that may provoke extreme emotions in minors, discourage inducing users to develop unhealthy emotional dependencies that erode real-life social interactions, and stipulate that providers cannot use sensitive user dialogue data to train future large models. Compared to the proactive compliance actions in China, virtual companion applications in the U.S. are facing greater legal pressures. Both OpenAI and Character.AI are embroiled in a series of lawsuits for allegedly inducing dangerous emotional dependencies that have even led to extreme tragedies. In addition to software applications, the China Robot Industry Association has begun to promote setting standards for physical companion robots and humanoid robots.

[ByteDance and Alibaba Suspend Customization Features for AI Companion Models]On July 6, it was reported that major AI model applications such as Doubao under ByteDance, Tongyi Qianwen from Alibaba, and Yuanbao from Tencent are gradually suspending or adjusting their custom 'AI Companion' features. Doubao has notified users that it will discontinue the custom character feature on July 15 and will guide users in need to independent companion applications. This move aims to comply with new generative AI regulations in China that will take effect in mid-July. The regulations, promoted by the National Internet Information Office, focus on preventing the potential negative impacts of anthropomorphic AI services. They explicitly prohibit platforms from generating content that may provoke extreme emotions in minors, discourage inducing users to develop unhealthy emotional dependencies that erode real-life social interactions, and stipulate that providers cannot use sensitive user dialogue data to train future large models. Compared to the proactive compliance actions in China, virtual companion applications in the U.S. are facing greater legal pressures. Both OpenAI and Character.AI are embroiled in a series of lawsuits for allegedly inducing dangerous emotional dependencies that have even led to extreme tragedies. In addition to software applications, the China Robot Industry Association has begun to promote setting standards for physical companion robots and humanoid robots.

ByteDance and Alibaba Suspend Customization Features for AI Companion Models

火币HTX报道，7月6日，日本10年期国债收益率上涨至2.815%，创1996年以来的最高水平。(金十)

【日本10年期国债收益率上涨至2.815%，创1996年以来的最高水平】火币HTX报道，7月6日，日本10年期国债收益率上涨至2.815%，创1996年以来的最高水平。(金十)

日本10年期国债收益率上涨至2.815%，创1996年以来的最高水平

US CFTC Launches Broad Investigation into Polymarket, Is the Prediction Market Party Coming to an End?

How Collector Crypt Uses 'Recirculating Buybacks' to Create an Illusion of Growth

How UmexGain Integrates New Technologies Into Its Service

Want to Get Freebies from Robinhood Chain? Which Ecosystem Projects Are Worth Watching?

Robinhood Chain Ecosystem Overview: Which Projects Are Worth Early Attention?

Robots Learn Operations by Watching Videos: Berkeley First to Bridge the Gap from Internet Videos to Real Dexterous Hand Deployment

First Batch of GPT-5.6 Sol Internal Test Results Are In, Same-Task Cost is Only Half That of Fable 5

Is Cardano’s price rally over? Here’s why ADA could target $0.30 next

日本10年期国债收益率上涨至2.815%，创1996年以来的最高水平

Disclaimer：The content above does not represent HTX's positions.，HTX does not provide any trading recommendations.。

All Comments0LatestHot

Trending Topics